• Subsidiary priced at 3x earnings of total market cap, with N.I mid teens CAGR.

• 2x separate offers at 2-2.5x higher than current valuation, rejected in last 12 months — clear M&A demand even at cyclical lows & unprofitability at the time.

• 0.06x P/S, deep fucking value.

• Proven, old management returned to clean up after ex CEO, signs of turnaround present.

• Global industry multi year cyclical lows, with initial global recovery in the last half year for the majority

The thesis is centered around these points, if this is not enough to illustrate the attractive opportunity, I suspect the rest of this write up is not worth reading.

LIQUIDITY: £14mil market cap & 80% shares outstanding owned by mostly a few institutions and a couple insiders.

I only post when I find truly asymmetric opportunities worth owning

Subscribe now

Structure

1. Reason For Opportunity

2. The Business

3. History

4. Management & Shareholders

5. Financials

6. Turnaround

7. Risks

8. Upside scenarios

REASON FOR OPPORTUNITY

There are a mix of factors which I believe to have caused the current pricing.

1. Size - £14mil nanocap, the smallest of the small & extremely illiquid.

2. Screens poorly due to:

   • Entire industry had been in multi year cyclical downturn which has just reverted as of Q3 2025

   • Return of old management and their turnaround results have yet to pull through as one-off paper losses, mask profitability

3. AIM - hated market

4. Good co. subsidiary buried among 15 others.

   ---

The business

Empresaria is a group of global recruiting businesses, who hire on behalf of other businesses, it’s split:
- 22% in permanent placements
- 51% in temporary/contract
- 27% in a single BPO offshore subsidiary, with 20 entities globally.

As per the visual above, revenues are mixed globally whilst UK & US make up 14% but worth mentioning the Offshore services (27%) is based out of primarily India, then Philippines, it actually mostly services its UK/US/Europe clientele.

The offshore BPO business (72% controlling interest) has rapidly scaled at a 32% CAGR from 2017 to 2023, with a slowdown to mid teens CAGR in most recent years to become 27% of Net Fee Income and the main profit driver.

Again referring to the visual staffing is mostly for skilled labour and split in 6x sectors:
Offshore services, Professional, IT, Healthcare, Construction & Commercial.

History

Founded in 1996 by Miles Hunt, ex-UK solicitor who scaled the business as a roll up where brands were bought and the local management, retained/owned 10%-40% of their subsidiary and remained directly responsible for their own P&L, the business was ran a decentralised model, very similar to the holding companies at Berkshire.

Following this model revenues scaled from £43.7mil in 2004 when listing on the AIM (originally at 65p) until 2011 where revenue grew to £208.9mil (25% CAGR)

Then in 2011 the founder retired and shortly after handed over to Joost Kruelen; ex-Managing Director of Vedior Professional Staffing Services, Netherlands, a subsidiary doing ~€500mil revenue at the time, before moving to Empresaria into Head of Asian Operations (Jan 2009) for 3 years before becoming CEO in 2012. Source

Joost Kruelen, continued to run the same decentralised serial acquirer playbook, which rolled up smaller recruiting business mostly in the £5-10mil range, with ownerships stakes no higher than 83% in any of the acquisitions to ensure local management remained aligned, for the 5x acquisitions made over his 7 years.

Joost later stepped down in 2018 as he semi-retired, moving back home to Netherlands and worked in purely non-exec director roles for the next few years.

Rhona Driggs joins first as COO before being promoted to CEO within 8 months in June 2019. Rhona had experience most recently as VP of a ~$1billion subsidiary, for Volt Information Services, a large recruiter, who opposite to Empresaria had ran a centralised business model rather than decentralised.

During her time she started shifting the business to a centralised model and split the business into ‘core’ & ‘non-core’ segments.

Straying away from the core business roots of running a decentralised operation where every subsidiary is accountable for their own results with every local manager having their own skin in game, with this structure starting to be replaced it contributed to the decaying performance, which also largely driven by a global industry wide cyclical downturn simultaneously starting 2023 where effectively all peers globally had declining earnings every year from this point up until very recently.

With the stock trading at depressed valuations and management at the time exploring options to sell underperforming subsidiaries, two offers had been received in May and July 2025 by 2x separate offers for 60p and 62p respectively whilst the stock was around 25p

Ultimately management at the time recommended the offer of 62p/share. This decision was opposed and overruled by the institutional shareholders who sided with industry legend/2004-2022 ex-Chairman/28% shareholder, Tony Martin who whilst not having an official role in the business today remains to have arguably the most significant influence in the business today, which works both ways.

Shortly followed was a whole reshuffle of the board, in late 2025, including herself (CEO), Chairman, CFO at the time, again led by Tony Martin. The new directors came on consisting of: Joost Kruelen (ex CEO, now Chairman); Spencer Wreford (ex Finance director, now CFO), Nigel Marsh in as CEO (previously 18 year CEO of Healthcare recruiting Medacs, which sold for £85mil in 2023)

Management & Shareholders

The management team consists of:

• Nigel Marsh, CEO

• Spencer Wreford, CFO

• Joost Kruelen, Chairman

• Vinod Tailor, Non-Exec Director

• Arun Shankardass, Non-Exec Director

• Eckhard Köhn, Non-Exec Director
  

Nigel Marsh, CEO, 63
Joined March 2026, as per most of the management following the overhaul, most are new members. Nigel has substantial experience, most importantly having been primarily CEO of Medacs Healthcare 2003-2021 (sold in 2023 for £85mil). Followed by mostly consulting/Non-exec roles for the next few years before joining Empresaria.

Needless to say he is a proven decade long operator of a same scale recruitment business. Interestingly he was one of the individuals behind one of the takeover bids, which will be addressed under risks.

Spencer Wreford, CFO, 54
Announced his return as CFO in March 2026, Spencer had actually worked mostly as Finance Director for Empresaria for around 9 years (2010-2019) during the same Era as Joost (now Chairman), when he had been CEO. During the end of his stint he filled in as interim CEO for 1year 2months.
Afterwards he spent 5 years as CFO for Penspen, an energy sector staffing services provider.
After in 2025 moved to Dataquest Group as CFO for only 11 months before taking the CFO role at Empresaria. The short stint is particularly telling as most of his roles have been 5+ years historically alongside his ties to the business and connection with Joost, it was highly probable he had been picked as the right man for the job.

Joost Kreulen, Chairman, 67
Rejoined in October 2025, as the first of the new exec team to kick off the shareholder backed management ‘overhaul’. Long industry history having senior roles in recruiting as managing directors for Staff Planing, Select Appointments UK and Vedior Professional Services (EUR 500mil annual revenue).

Head of Asian Operations (Jan 2009) for 3 years before becoming CEO in 2012 till 2018.
During his tenure the business had consistently growing topline and maintained steady profitability, which could not have been said for Rhona, who came after Joost.

He departed in 2018, to Semi-retire, evidenced by his move back to his home country Netherlands and only taking on non-exec roles for the next few years.
His Linkedin posts get a sense for his vision and management style, here are a few.

1) Love this one, had to include it!

2) Values traditional in person connections.

3) Again, reinforcing people are core of strategy.

4) Focus on higher margin ‘perfect storms’

Vinod Tailor, Non-Exec Director, 68
Joined October 2025 as the ‘Capital markets’ expert, 40+ years experience in financial services for Elara Capital, Merrill Lynch International, Citibank.

Arun Shankardass, Non-Exec Director, 79
Joined October 2025, finance background, historically primarily worked as Chartered accountant or Corporate treasurer. Previous roles at Yves Saint Laurent, IC Gas, Aviva, and corporate treasury at Lex Services Group.

Eckhard Köhn, Non-Exec Director, 56
Joined October 2025, previous experience as CEO of JobValley, Germany's largest student, workforce-as-a-service provider (EUR 100mil revenue in 2020). Important considering historically Germany has been a struggling region universally across recruitment businesses.

As you can see the Non-Exec directors were all effectively put in at the same time, this was done following shareholder approval and led by largest shareholder and exChairman; Tony Martin.

Management Ownership:
Effectively management does not own any stock, with only Spencer Wreford owning a token 15,000 shares valued at £3,000.

I would prefer if management had large skin in the game, however I am equally well aware that we have proven directors who have shown to historically perform well, even with this in mind.

Shareholders:

As visible based on the most report, institutions/insiders effectively make up 81.07% of the overall register. Making the business illiquid even for microcaps.

Tony Martin - 27.93%
Largest shareholder and in my opinion, the most influential person at Empresaria and one of the most influential within industry as someone who:
- Ran half of Adia S.A at the time the worlds second largest staffing group at the time
- 1992-1999 ran Selection Appointment Holdings plc as Chairman and CEO, growing and then selling to Vedior for $1.8Billion in 1999
- At Vedior, he got promoted to CEO and Chairman in 2000 from which till 2004 before retiring, shortly after sold to Randstad for €3.51 billion in 2006.
(The joint entity was the worlds second largest recruiter).
- Had also been the Chairman of Federation of Recruitment and Employment Services, in the UK.

The remaining are primarily split between concealed small funds as well as the Trinity Bridge & UBS position most likely being a nominee for a UHNW client or institution. It is worth pointing out that a good number of these shareholders are Dutch entities.

Previously an agreement had been signed amongst them to hand over voting rights on a prior vote at the time; H M van Heijst controlled its 17.89% and Beleggingsclub ‘t Stockpaert — 3.51% and Stichting Value Partners Family Office — ~7.7%.

The overarching theme here is the majority of the register is owned by sophisticated institutional/insider owners.

Financials

Net Loss Breakdown

On paper Empresaria had another significant loss this year at (£7.6mil).

1. Impairment of Goodwill,(£5.3mil) this one off impairment has risen significantly, whilst it is important to note that the reason behind such a large impairment is likely due to the continued underperforming German subsidiary which like most recruiting businesses in Germany has been struggling and remains with a £12.8mil goodwill value associated with the subsidiary itself and only £21.6mil total goodwill value remaining for the business itself.
   Frankly I also see the new management coming in this year and doing the often used playbook of creating massive write-offs and setting expectations low, to set themselves up for success later on as factor behind the significant jump.

   I do not see £5.3 million write-offs likely to be a continued repeatable ‘cost’ in future years.

2. Ammortisation of Intagible assets (£1.1mil), again this is a legacy cost associated with intangible asset values tied to customer relationships and trade names which are for acquisitons mostly having last occurred in mid 2010s. The gross initial amount being £20..7mil with only £4.3mil remaining which has yet to be depreciated, providing only a few more years of runway.
   Again of the (£1.1mil) annual charge the only business cost I see is the (£0.1mil) annual software depreciation.
   Making the remaining £1mil annually a non business cost, particularly as this reserve will empty in the next 3/4 years.

3. Exceptional items, (£2mil)
   The below breakdown has been reported in the financials.

   

   It is confirmed on Page 19, in the FY 2025 report that:
   Board costs, (£0.9mil) ”£0.9m for the termination costs for the departure of the former executives and reconstitution of the Board in October 2025”

   Legal costs, (£0.7mil) “£0.7m for legal and professional costs incurred in respect of the aborted takeover offer”

   
   Of the (£2mil) one-off cost, I do not see the board changing every year, nor do I see the business rejecting takeover bid offers every year, these to me are truly low probability of reoccurring annually. Meaning next year I expect a (£1.6mil) reduction in losses associated with these.

In total the Empresaria lost, (£7.6mil) on paper for FY 2025. When adjusting to remove one off and non-cash tangible depreciation costs.
(£7.6mil) + £5.3mil + £1mil + £0.9mil + £0.7mil = +£0.3mil Adjusted Net Income.

Whilst this does not quite match managements adjusted profit profit before tax of £4mil figure, to me it highlights that the business has returned to profitability/is much closer to profitability than the financials show at a glance.

Cashflow

Empresaria had operating cashflow jump from £1.4mil to £7.7mil annually! (whilst revenue continued to decline (3%)YoY)
However the true figure is more like £5.9mil (when you adjust for non-controlling interests (most of which is the 28% stake in the IMS offshore subsidiary which EMR does not own)

As shown in the statement the majority of this £5.9mil cash has gone towards paying off debt, in this case mostly lease obligations and totalling (£5.6mil) net paid towards finance costs.

Balance Sheet

Trade receivables of £38.2mil are significant, however this is a part of the business structure particularly as a large amount of its business lies in temporary contract placements where Empresaria first pays the contractors wages before being paid by the company 30-60 days later.
”Average debtor days for the Group in 2025 reduced to 38 days (2024: 39)” which above peers who generally run between 40-50 days and higher depending on the type of recruiting business model.
”The income statement includes a charge of £0.5m (2024: £3.2m) in respect of impairment losses on trade receivables”. Trade receivable losses normalised for context 2022: £nil; 2023 (£0.3mil); 2024 (£3.2mil); 2025 (£0.5mil)

Net Liabilities
With current assets of £58.2mil and total liabilities of £75.6mil this business has net liabilities (when only counting current assets) of (£17.4 million) this is a large amount of for the scale of the business and the reason behind why operating cash is being used to pay off debts.

Of which £71.7million is due as current liabilities mostly split between:

• £28.9mil payables

• £34.7mil borrowings

• £8.1mil lease and deferred tax liability

Borrowings is the largest segment which is split in the following:

Mostly concentrated in the UK holding company, of which the £15mil revolving credit facility, since the FY 25 annual report, it was extended 13 months to October 2027, previously meant to expire September 2026.

There is some room to further increase the borrowing by a further £5.4million, as well potentially drawing on invoice financing in a significant way, however that would not be ideal, given it is more costly than traditional borrowing.

Non-Current Assets
The £37.6mil Non-current assets, are mostly made up of Goodwill and Intangibles (trademarks and relationships), little tangible assets which is expected for an asset light, people heavy business.

Within non-current assets the main items of tangible value would be the remaining £6.4mil right-of-use assets, mostly leased office space and to some extent, the £3.9mil tax deferred assets

IMS Subsidiary Breakdown

Results here have been consolidated to account for Empresaria’s 72% ownership of IMS group.

Revenue growth in recent years has slowed with growth YoY of 6.7% and Net Income growth of 17.4% YoY. Keeping in mind these results are coming off a year where most peers have been in decline.

Strong individual balance sheet, however as later mentioned under ‘risks’ temporarily whilst the UK subsidiary remains unprofitable, any funds moved away from India and back the UK would effectively be taxed at a 60.5% rate.

This is the crown jewel business hidden in Empresaria:
- Historic 32% CAGR from 2017 to 2023. Slowed down since to midteens Net income growth rate.
- 18.8% net income margin, high relative to peers.
- Structural growing BPO offshore market, a ~9.8% CAGR projected for global BPO till 2030.

Turnaround

For the last few years earnings have been declining there have been two elements at play here, both of which I believe to not be fully reflected in the current numbers.

1. Management Turnaround

2. Global shift away from Cyclical Downturn

Management Turnaround

Previously from the year 2018 the business had been ran by the then old management team using a decentralised model, the exact same model used since its founding in 1996.
The main advantages to this model meant every subsidiary had both autonomy and responsibility for their own P&L, paired with most local managers having sizeable amounts of their personal net worth tied to the ownership of the subsidiary they ran.

One of the key factors, that I can see pointing towards declining performance had been the shift away from this very structure into a centralised model which Rhona Driggs, previously VP at Volt Information Sciences (who also had a centralised structure).

This revised structure meant less:
- performance based incentives
- Direct P&L responsibilities
- Less dependence on the boots on the ground, replaced by decision making from isolated headquarters.

As well as splitting the business into ‘core’ and ‘non-core’ segments with an aim to dispose of non-core and centralise under a core brand. Management attempted to fix a model which had not been broken at the time and failed.

Following majority shareholder alignment on the vote led by Tony Martin, in October 2025, a full board refresh occured.
As outlined under ‘Management & Shareholders’, the Chairman alongside 3x new Non-Exec directors had been brought in. Chairman, Joost, had later brought in ex-colleague, Spencer Wreford, then Finance director of Empresaria, now CFO.

As well as CEO, Nigel Marsh, long time Medacs Healthcare CEO. Whilst there has been no letter from the CEO as of yet the Chairman letter to shareholders is telling.

Key points being:
- Focus on return to decentralised leadership “ Back to our founding DNA”
- “Moving away from “Core” and “Non-core” classifications. This structure restores local accountability” bringing back individual, strengthening individual P&L responsibility.
- “focus shifts decisively from “growth at all costs” to Group’s balanced, profitable growth” focus back to profitability.

Keeping in mind this is the same person who grew and ran the business profitably as CEO, from 2012-2018, without a single losing year.

It is still early however a marked positive improvement in financial results can be seen considering the increase in operating cashflow from £1.4mil to £7.7mil as well as operating loss decreasing from (£3.6mil) to (£2.7mil), particularly as net fee income still continued to fall (4%) YoY.

2.Global shift away from Cyclical Downturn

Generally the majority of recruitment companies outside of a few small pockets (specialised tech / Japan) have faced multi year declines in Net Fee Income as part of a cylical downturn originally starting in late 2022/early 2023

Recently the industry has started turning back to growth roughly in the last half year, with some as early as Q2 2025.

I have gone through every single recruitment company globally I could find and reviewed their most recent earnings transcripts as well as general earnings to get a pulse on the industry, results below:

Globally Diversified Majors:

Randstad - Q2 2026 “overall revenue growth is picking up, delivering 1.9% organic growth for the quarter”. Previously had consecutive declines in revenue since 2022

Adecco - Q2 2026 “We have seen this progress across several countries, including the U.S., Spain, and APAC, where permanent placement gross profit grew positively in Q2” Returned to growth from Q3 2025, prior to this the business had declining sales since Q3 2023

ManpowerGroup - Q2 2026 “moved from stabilization into a recovery”. Sales declines started in Q2 2022 and only reverted back to growth in Q3 2025.

UK Professional permanent specialists:

Hays - 07/10/2026 “We currently expect FY 2026 pre-exceptional operating profit will be at the top of the consensus range, following a strong return to year-on-year profit growth”

Whilst revenue declines have continued, YOY decreases have dropped to (3%) whereas year prior business shrunk (5%) and (8%) year before.

Michael Page - Since declines starting H2 2023, the business turned back to growth in H1 2026 (though this was marginal 0.13% growth, effectively flat)

Robert Walters - H1 2026 back to 7% growth, similarly this business had been in consistent revenue declines since 2H 2023.

SThree - 1H 2026 “Trading momentum improved through the first half despite continued macroeconomic and geopolitical uncertainty” Whilst the declines in sales continued, there was a slowdown, from a (7%) decline versus year prior (15%) decline.
Again much like peers sales had peaked in 1H 2023.

US professional & tech staffing

Robert Half - Q2 2025 “Hiring demand continues to improve, market conditions are increasingly more supportive”
To no surprise sales had also peaked mid 2023 with revenue declines falling to only (2%) YOY, prior year (7%) declines.

Kforce - Q2 2026 “Overall revenues positively inflected in the first quarter of 2026, meaningfully expanded in the second quarter, and our guidance for the third quarter contemplates continued sequential improvement”. The business returned to growth in Q1 2026 with revenues having initially peaked in Q4 2022

Everforth - Q2 2026 “We saw a meaningful acceleration in the business during the quarter.”

Korn Ferry - Q4 2026 “we grew for the fifth consecutive quarter”

Engineering / energy / technical specialists

Brunel International - 1H 2026 "Permanent recruitment delivered organic revenue growth for the second consecutive quarter."

Gattaca - 1H 2026 - “the majority of our sectors experiencing year on year growth”

Blue-collar / industrial

Staffline - the business returned to growth in 2024.

TrueBlue - returned to growth in Q2 2025; much like its blue collar peer, recovery has occurred quicker than other types of recruitment professionals.

Summary

As you can see by the above commentary the majority of the industry has shifted back to growth and improving profitability, whilst this is still early for a lot of these peers relative to peak earnings formed mostly in 2022/2023.
Empresaria itself has confirmed that “net fee income for H1 FY 2026 up approximately 5%”, growing for the first time since 2022.
As well as “now expects adjusted profit before tax for the year ending 31 December 2026 to be at least £5.2m” versus FY 2025 adjusted profit before tax of £4.0million.

This has even started to be priced into the valuations of companies, with the majority of these businesses up 50%-100% in the last 1.5month from multi year lows.

Stock Performance

Empresaria on the other hand is only up 20% during this time (slightly more since the screenshot), I suspect due to the business having low visibility due to having only 1x sell side analyst and a £14mil market cap.

Risks

Debt and balance sheet refinancing risk
As addressed under ‘Financials’ the business has a net liability position of (£17.4 million), when counting only current assets, this is a large amount of debt.

Whilst the business has returned cashflow positive recently and extended its £15mil Revolving Credit Facility to October 2027, there is no guarantee that Empresaria will receive sufficient or similar funding terms as previously.

Lending criteria could tighten, e.g should a recession/black swan event occur or business performance falls off.
In fairness the business is already cashflow positive, with still room for up to further £5.4mil in borrowings and optionality to use invoice financing if needed, but these are ultimately short term fixes and would rely on the business sustaining profitability.

Worst case there are also some attractive subsidiaries which could be offloaded to secure cash. It is also worth mentioning management are still planning to strategically dispose of some of the subsidiaries, my sense is this is to remove low margin players, whilst paying off debt.

Performance
It’s not clear if we are yet out of the ‘downturn’ only that we are seeing signs of recovery and return to growth across most players in recruiting, there have been instances where the market has reverted back to a downturn shortly after.

Or its very possible, that CEO, Nigel who has historically focused more exclusively on healthcare recruiting may not be able to ‘turnaround’ loss making subsidiaries.

Double Taxation
The IMS subsidiary which currently generates £5mil+ net income annually, is being double taxed due to the holding structure on any funds coming back to the UK, should the UK subsidiary not return to profitability (where its profits can be offset) the business would remain being taxed at a rate of 60.5% longer term.

This is the reason behind a large amount of cash being sat in the Indian IMS subsidiary, on the assumption the UK returns to profitability, funds can later be moved at a more tax-efficient rate.

Takeout Offer
It would not be surprising for me to see management (particularly those who were involved in the original bid offer) to look to attempt another buyout on the business.

Whilst I do not suspect that the sophisticated institutional shareholder base will easily get taken advantage of in terms of pricing, it is possible for a deal to go through at an unfavourable price to shareholders. With a 60p and 62p precedent it is likely to not be too much lower, however this again is not guaranteed and only a possibility, e.g should a recession occur, this can be lower.

Additionally worth noting that AIM has relatively weak buyout rules which often can favour the purchaser, I have seen an instance of a company being bought out at only a 2% premium to the most recently traded price, which had been down 60% from all time highs, though it did follow with lawsuits.

Liquidity
Given this is a £14mil nano/micro cap AND 81% is owned by a few large generally stagnant owners, the business has low liquidity even relative to other companies of its small scale.

Exiting a meaningful amount in any timely manner, may not be possible or very costly if done.

Upside scenarios

There are multiple ways to value the business here, we will go through:
- M&A
- IMS subsidiary valuation

M&A Demand

1. Planmatics Limited — offered on 7 May 2025 OFFER: 10p per share cash + 50p per share unsecured 3 year loan notes (60p total)

• Individuals: Peter Gregory, Nigel Marsh (current CEO) and Ashok Vithlani (co-owner of non-held 28% stake of BPO business)

2. Legacy UK Holdings Limited — offered on 30 July 2025 OFFER: 62p per share cash. (148% premium to pre-bid price)

• Tied to the Khanna family, New Delhi–based hoteliers; Yajush Khanna is CEO of Jukaso Hotels Palaces and Resorts and Legacy Mining.

• Management team at the time recommended this offer, however major shareholders felt this undervalued the business and instead opted to bring in old mgmt team to run the business.

• Deadline: extended 27 Aug → 24 Sep 2025 → 22 Oct 2025

• Outcome: withdrew 22 Oct 2025 — after the 15 October board changes, the purchaser withdrew seeing it no longer had any support for its offer to go through.

As of writing the stock price has increased to 30p/share, whilst this is higher then when I initially started writing, it is still visible that 2x separate entities believe the business is worth double if not more than the current valuation today.

Bearing in mind these offers had been presented when the industry had still remained in trough demand and had not yet started recovering. The business and market is in a better position today and again these offers were not sufficient for shareholders.

IMS subsidiary valuation

Generating £5.4mil profit on a consolidated basis, whilst compounding earnings at midteens this business alone is worth atleast 10x earnings
(Peers in India can often trade much richer).
£5.4mil x10 = £54mil - (£17.4mil) net liabilities = £36.6million market cap

This factors in the price of the IMS subsidiary, net of the outstanding company debts meaning the subsidiary alone could be worth double the market cap today.
This assumes the remaining ~70% of Empresaria is valued at 0.

Earnings Multiples

Normally I am a fan of using earnings multiples against peers as a method of checking, however I believe backward looking financials have yet to have caught up with this business, therefore none have been provided.

If I had to summarise my thesis in one a couple sentences it would be:

Massively overqualified management team, who have successfully ran the business previously, now returned. 1x subsidiary alone generates near half the market cap in annual net income, whilst the industry is recovering from multi-year cyclical lows and remaining subsidiaries turn to profitability.

Strong M&A even during industry trough, where buyout upside is multiples of current market cap.

Any questions please drop me a comment, its how I learn more about the business too e.g I wasnt aware about the double taxation, till it was mentioned to me.

My next deep dive will only be once I find the next unreasonable opportunity
Thank you for reading,

Recently Ive seen a few interesting ideas recently, but upon looking further into the annual report, business model, etc, the thesis starts falling apart, would only be deploying capital when a worthwhile pitch comes through

Disclaimer: All content is provided for informational purposes only and does not constitute financial, investment, or trading advice or any form of professional guidance.

Newtek

02/12/2025 — $10.87 │ Now — $12.68
Core Thesis: Private credit now turned bank, 4.83x P/E, 17%+ Revenue CAGR, Funding mix shifting to deposits away from expensive private capital (huge interest margin expansion underway), Counter-Cyclical growth (faster growth in recessions). Free deep dive

Following earnings, my key takeaways:

Interest Rates on Savings — Storing Future Growth
Exceptional deposit growth Q1 ‘25 - $966mil → Q1 ‘26 $1.859bil +92% Growth
Vastly exceeded the initial 30% deposit growth projected

Newtek has always taken the approach of offering among the most competitive rates to retain & win customers, especially because it could afford to do so with its high yielding business loans.

High Yield Saving Account (HYSA) interest rate lowered from (July ‘24) 5.25% → 4.2% (May ‘26) with business saver having been lowered to 3.5% intentionally less competitive than some competitors providing 4%.
Whilst the fall is mostly due to interest rates having fallen, Newtek could have very easily, lessened rate cuts, but intentionally opted to (important part) ‘control growth’ so that the business can sustainably scale, they also now have a waiting list for their deposit accounts, which makes me optimistic on the ability to continuing compounding long term.

Q1 2026: 78% of deposits were FDIC insured (funds totalling $250k or less) versus 79% Q1 2025. Having well diversified source of deposits, again shows the quality of customers which Newtek has rather than chasing growth at all costs and its still putting up these numbers.

Funding Shift
This time last year the business was funded ~50/50 between Deposits & Credit facilities/Investor funded notes. Now this is ~77/23 split as Credit facilities have been completely paid off and Investor funded notes are planned to be paid off when the expiration date becomes due. Huge for profitability as Newtek only pays ~4% on deposits whereas Credit facilities cost ~7.6%.

I initially anticipated deposits to be 80% of funding in 2028, whereas we are effectively at this target in only Q1 2026, this is largely the reason behind why Newtek has been able to grow profitability even though loan growth has been flat.

AI implementation
The business has been actively implementing AI features to name a few:
- Upselling with automated Property&Casualty insurance for businesses
- Some SBA loans are now processed with the use of AI and even where it is manually underwritten, documents are still cross checked by AI as well as other admin features.

“slower start to the year due to implementing process enhancements and staff training, we were encouraged to see the number of units originated surge in the second half of the quarter following the launch of the Newtek [SBA] Seven Day Business Loan”

Improved Efficiency
Huge efficiency gains in Newtek bank structure, efficiency ratio:
Q1 2025, 48.2% → Q1 2026, 40%

This has largely been driven by AI, as staff count has remained relatively flat whilst growing deposits +92%. Additionally there are decreased costs associated with ‘professional services’ seems to be driven by no longer needing to invest as heavily into the new ‘bank structure’.

Regulation Growth Headwinds
SBA loan volume demand has dropped an estimated 10/15% as 100% US ownership rules have been enforced, this has historically been anywhere from 50%-65% of Newtek’s total loan volume, this is a short term issue and 1 year out from now will be normalised.
Loan origination of $1.42bil in FY 25, whereas FY 26 is guided at $1.8bil +26% growth.

Profit Guidance
Analyst projected 2027 EPS growth of $2.40 → $2.47
Management confirmed 2027 EPS guide of $2.40 → $2.60

However would not be surprised if this is conservative, considering the track record of meeting or beating EPS guidance.

Thoughts
The business remains at a compelling valuation (5.6x P/E), with management that continues to impress with their execution, whilst flat loan growth was disappointing there are legitimate reasons behind this with guided growth.

Though my largest holding, it is a rare position I feel myself wanting to add more to over time, not less.

Kaspi

19/11/2025 — $72.80 │ Now — $87.51
Core Thesis: 5.7x forward P/E, 20%+ CAGR, Monopoly with one of the strongest MOATs globally, operating as a true ‘SuperApp Ecosystem’ in Kazakhstan with a recent Türkiye acquisition (Hepsiburada) Free deep dive

Hepsiburada
The 2nd largest Turkish e-commerce platform, Kaspi in 29/1/25 took a 65.4% position for $1.12billion. Since then Kaspi has been actively purchasing shares in the open market & through capital raise, now owning 85.66% .

Ownership Changes
→ 20year shareholder, private equity fund Baring Vostok, has been selling of position, still owning 22.3% of Kaspi.
→ Tencent the 1 billion+ user ecosystem giant, has taken a ~5% stake into Kaspi
→ CEO Mikhail invested further, buying a further 0.5%, now owning 23%

Frequency of Transaction (yearly)
Following the earnings, one thing Ive come to appreciate more with Hepsiburada acquisition, is the low hanging fruit from increasing customer engagement/frequency of purchase.
Hepsiburada currently makes up ~16% of Kaspi revenue, I see significant room to expand the business from its existing customer base by increasing the stickiness of the product & overall frequency of purchasing for context:
(Annual purchases per Customer)
7x/Yr Hepsiburada Türkiye → 27x/Yr Kaspi Kazakhstan → 40x/Yr per U.S Avg

I’m not suggesting that Türkiye will match the US, but I think there is possibility for this to become a multiplier of Hepsiburada’s current scale as the gap narrows over time as Türkiye population increase frequency online purchases and Kaspi improves the stickiness and value provided in its services.

Growth
Frankly this has become very messy to provide direct comparisons with:
Hepsiburada acquisition, currency fluctuations, fintech business model shift, increases in interest rates.

Q1 Revenue in KZT was up 31.5% on last years Q1 revenue, however again this takes into account an extra month of Hepsi revenue and other factors

31.5% KZT growth was split into +7%, Payments, +49%, Marketplace & +25%, Fintech growth on YOY comparisons.

Payments growth was unimpressive (potentially flat, in real terms) as the business grew share but cannibalised its own fees as a larger portion of its transactions were on the lower 0.95% fee, instead of ~1.5–2%+ because B2B payments and Kaspi QR grew disproportionately fast, both by design have lower take rates than standard consumer transactions. Im not concerned here considering that fast, low fee transactions is one of the main way customers are brought & retained in the ecosystem

Marketplace obviously had a boost from Hepsi, but a key metric which stood out was that the quarterly order frequency per quarter (for Kazakhstan) which rose from 10.4→15/quarter. +44% jump in frequency which again goes back to my point of the significant room the business has to grow from order frequency alone, outside of potential to increase:
- average order value
- number of customers
- further cross selling

Fintech grew at a nice 25% which included a revenue shift to interest income from fee-based BNPL’s because Kaspi is pivoting to focus on extended average loans, of which average duration increased from 7→9.3 months, however total newly originated volume was (-2%), which was part of the tradeoff for higher lifetime yield.

However it is worth noting that new user growth was only ~6%-7% increase however considered Kaspi, is in the rare spot of actually nearing physically maxing out the country (in my deep dive I estimated a real Kazakhstan market share of ~60-70%) I expect the growth to continue from population increases & further digitisation but with the majority of new user growth coming from Türkiye as it becomes a larger segment.

Catalyst?
Initially cutting the dividend to fund the acquisition, Kaspi has returned to offering the dividend at a 60%+ pay-out ratio, which the board has recommended. This works out to an 8.4% annual dividend yield.

Speaking of Catalysts Mohnish Pabrai has also taken a 4%+ position in his Wagons fund portfolio. Listening to his Wagons fund update, the essence of his thesis is: World class management, low valuation and lower associated risk, large upside from Türkiye expansion (multiple times larger than Kazakh market), whilst still dominating in Kazakhstan.

Thoughts: Solid earnings, the business continues to execute, Im already significantly concentrated with this being my 2nd largest position, do not see myself selling nor adding.

Next is a position I’ve not talked about before in depth, this is a historic position prior to starting publishing on Substack and currently my smallest, after having taken profits, but still a meaningful size as it is 1 of 3 positions held in the portfolio.

Ring Energy

17/7/2025 — $0.77 │ Now — $1.36
Core Thesis: Oil & gas producer in the Permain basin. 35% shareholder just finished its forced sale (~30%+ of daily selling pressure vanishes overnight). Selling as part of transition to ESG fund, Trading at 12.2% of PV10 using conservative $62 Oil assumptions (Peers trading at ~60% of PV10), Mid-teen ROIC rollup by implementing new drilling methods in old oil fields left for dead

CFO hire
Hired Sonu Johl, 20 Years experience, Energy investment banking background, advised on acquisitions in the Permian basin, co-Head for Raymond James and UBS for E&P.

No operational CFO experience but direct M&A experience showing the aim the business is looking to focus on, great however my hunch is that the prior CFOs had left over concerns with high debt levels and frankly what I speculate (my own opinion, NOT factual) is McKinney’s aggressive expansion plans, for example, most recent acquisition, which was excellent for ROIC, but horrible from a risk management perspective, considering their heavy debt loads and uncertain declining & deadly low oil prices at the time.

My concern with the hire, is that McKinney (CEO) has brought in someone who ‘fits the style’ of aggressive expansion, making him prone to similar flaws and carries less of the CFO cautionary burden which many conventional CFOs would have gained in their more traditional risk management orientated roles.

This is not to discount the experience of participating in numerous acquisitions and will be highly valuable for future acquisition & hopefully even more so for the pattern recognition of those who crashed and burned after acquisitions.

Debt Service
Debt service pay down had actually been occurring over the last ~1.5 year (unlike years prior, this had only been lip service), though this has happened arguably largely out of necessity.

Conversely, this quarter, using FCF to pay down debt has been put on hold to focus on early investments to ramp production, given current oil prices, there is a large buffer as far from its break even prices.

Oil Prices & Collar Hedge Contracts
Based on daily production, ~72% of oil contracts have collar hedges where for those barrels of oil, $73.27 is the ceiling, which limits upside of current oil prices until YE 2026, though there would still be a blended realised price of ~$79/barrel (assuming the rest is sold at the current $99/barrel).

Cash Dilution
Recently REI has done a secondary offering on May 12, 2026 for ~21% dilution and $56M net cash ($60M gross) & besides crashing the stock price by ~23% in a day, the proceeds will be used to repay borrowings under its senior secured revolving credit facility (which had $426M outstanding, so ~13% of it)

Again this evidences my thinking behind the less than satisfactory over aggression at times which has led to Ring Energy needing to dilute shareholders. Lets take the most recent 2025, Limerock acquisition, which occurred in a period of already low oil prices, not like the management had been taken by surprise that their cashflow is becoming strained. So what were the (approx.) results of this ‘accretive’ acquisition:

Acquisition cost (inc Dilution): $96million
Oil production increase: +14% (~1,840 Bo/d on ~13,000 Bo/d base)
Proved reserves increase: +9% (~12 MMBoe on 134.2 MMBoe base)
Estimated 2026 oil output: ~12% (declined from ~14%)

Now you’ve raised, only ~58% amount of cash needed to pay for the acquisition: Increasing production ~14% & Diluting shareholders by 21%
Sure, there were real synergies from the site being located next to an existing one but surely not enough to consider the next dilution effect & further levering an already highly leveraged business.

Thoughts
I came into the position, open to the risks of McKinney’s operating style which includes using fairly aggressive debt to continuously fund acquisitions at a fast rate (though I did not see another acquisition coming at the already high debt levels)
The price was low enough where it still made sense with the risks, due to the value of the assets.

My position size at entry was approximately half my usual, additionally this is why I exited roughly half my position, with solid gains, given there is balance sheet risk.

Given high oil prices, I do not see Ring struggling with debt for time being and that is why I will continue to hold the position, with the intention to cash out my gains at some point given my diminishing conviction in management execution

As always happy to answer any questions you have and let me know if its worth me doing more of these earnings update style posts

Thanks,

Quick Pitch:

• 3.84x P/E, 2.9x forward P/E

• Revenue 10%+ CAGR expected with 38% growth secured, driven by Mexico contract wins in 2H 2025 alone.

• Bottom ~10%-15% on cost curve
  (30.5% gross margins VS peer 12%-16% gross margins)

• Founder led, 70% Insider and Institution ownership.

1.What CT Automotive Does:

CT Automotive (AIM: CTA) is a £20 million market cap Tier-2 automotive supplier, in plastic injection moulded parts. The business manufactures non-strategic, yet still essential components — air vents, decorative trim, lighting housings, cupholders — for major OEMs including Nissan, Ford, GM, Volkswagen Audi Group, and what appears to be Tesla (referred to only as a “US-based major EV OEM” under NDA).

Founded in 2000 by Simon Phillips, a first-generation immigrant who built his first business importing electronics while at university, later working under a founder at a plastic injection moulding specialist business, fast forward to starting CT Automotive, post sale of an earlier venture.
Simon moved early to China when globalisation was still in its early days. The business had a structural cost advantage: Western engineering and standards & Shenzhen’s manufacturing ecosystem.

Phillips later sold the business via MBO in 2016 as he wanted to cash out and exit as priorities in life changed, to focus on raising his children but was brought back in 2020 when COVID and management missteps nearly sank it. He took CT public on AIM in 2021 to raise capital to deleverage the balance sheet.

Today, CT operates facilities in primarily China (Shenzhen and Ganzhou), Mexico, with supporting facilities in Czech Republic, and Türkiye. The Mexico facility, which began in 2022, now ramping and expected to be a significant contributor by 2H ‘26, I estimate 20%+ based on contract wins. Representing a strategic pivot away from previous ~70% China-centric production amid anticipated escalating tariff risks.

2. Why this opportunity exists?

No Discovery:

• £20m market cap — zero analyst coverage, given small size & limited float.

• Major Shareholder Selling Pressure - Otus Capital selling 18% stake over 18 months, now only 9% left, forced selling creates an extra 20-35% selling pressure of average daily volume. Forcing price down over time.

• AIM liquidity trap — too small for new institutions. Investing in the UK markets has fallen out of favour given poor historic returns and locals in the main preferring to stick funds into a tax-free Cash ISA at 4%-5% yields. Where retail investors remain, unless there is a dividend, they usually are not interested.

Screens Poorly:

• 3.15% five-year lumpy revenue growth — looks mediocre at a quick glance:

  • 2020: China COVID lockdowns

  • 2022: 2nd Wave Covid manufacturing extended shutdowns (in China)

  • 2023: OEM overproduction (compensating for 2022, which was largely missed)

  • 2024: Destocking cycle

  • 2025: Tariff-driven pauses on spending

• Screens can’t distinguish temporary disruption from permanent impairment — last five years non-representative of next five. Most unaware, OEM volumes haven’t fully recovered since pre-Covid.

• Hidden Strengths — margin resilience as revenue falls & customer retention.

3. The Apparent Bull Case

Revenue Growth and Recent Contract Wins

In 1H 2025 alone, CT secured contracts worth $37 million annually — representing 38% of FY2024’s $120 million total revenue. Four of these eight contracts were “programs won from competitors relocating production to our Mexico facility.”
Management has said they are preparing for a “step-change in volumes forecast for 2026.”

New Mexico facility is purpose-built for automation and robotics, the business set to benefit immensely from their proactive approach as “all time high RFQ” interest is driven by OEMs looking to shift production out of China. Facility expected to be at significant capacity by 2H 2026 (est. 20%+ based on contract wins), which reduces China dependency & would see margins and free cashflow increase significantly as the facility shifts from being a drag to a large contributor while maintaining CT’s low-cost position.

Unreasonable Valuation

3.84x P/E and ~2.9x fwd P/E, CT has room to increase its multiple 2x-3x to become inline with peers, who typically trade at 7-11x earnings. Adjusted free cash flow of £2.794mil (excluding, one-off’s — which I will breakdown in Full Deep Dive, if this becomes a position I hold). Versus £20mil market cap equals, ~7.16x FCF multiple — this is compressed, considering volumes for the industry are below pre-COVID levels.

Potential for Strong Balance sheet: Net current assets (current assets minus total liabilities) are £16.3 million, 81.5% of market cap. This is on paper as dependant on receivables being realised (addressed later).

Low-Cost Producer Moat

This is a MOAT for CT, in a commoditised price driven industry.
Estimated bottom 10-15% on cost curve for the industry

Peer gross margin benchmarks (Tier-1 suppliers with scale advantages):

Margin expansion continues to occur even while revenue fell 16% in 2025 — gross margins improved from 21.6% (2023) to 27.6% (2024) to 30.5% (1H 2025).
Automation efficiency in action, rather than use of buzzwords. Recent Türkiye job cuts due to automation redundancies provide tangible evidence.

Disciplined investment approach for automation ROI:
Every project must deliver payback within 1-2 years, as the lowest acceptable payback period compared to industry-standard 3-5 year payback periods.
Alongside faster execution due to minimal legacy drag from newer facilities and automation-first culture (actively pushed to implement unlike competitors) creates sustainable cost advantage, as it is built into the way the business operates.

Customer Lock-In and Product Positioning

Component contracts typically last 5-7 years (the life of a vehicle platform).
Heavy Switching costs: Switching requires re-tooling, re-validation testing, production line downtime, and quality risk — does not make sense for non-strategic, low-value parts. OEMs don’t bother unless there’s a major price difference or quality failure.

CT’s parts are platform-agnostic: air vents, lighting housings, decorative trim, and cupholders are needed in ICE, EV, hybrid, autonomous.
Even Tesla, which is heavily vertically integrated, appears to outsource these components to CT (again assumed based on ‘Major US based EV manufacturer) because they’re labour & tooling-heavy with near zero strategic value.
Tesla focuses in-house attention on AI, batteries, and software; outsourcing to the low-cost parts makes sense.

Founder Alignment and Forward-Thinking Strategy

Simon Phillips owns ~25% of the business. He has the founders ‘it’ factor — navigated the 2016-2019 MBO disaster, COVID shutdowns, and brought the company back from near-death.

His strategic foresight is proven: he built out the Czech facility to reduce China dependency and started Mexico facility early in 2022, anticipating US-China trade escalation since 2022.

Management has a ruthless focus on the 5-year vision to become the
“lowest cost, most competitive player with fastest and most advanced implementation of AI.” They’re leveraging decade plus experience in Shenzhen (global robotics hub) and encouraging company-wide LLM adoption and robotics implementation.

Tariff Tailwind

Four of eight 1H 2025 contract wins came from competitors relocating to Mexico. As the lowest-cost producer with large Mexico capacity, CT is positioned to benefit as OEMs prioritise cost reduction to offset tariff impacts, particularly as a low cost player.

Why Self Driving could be a Tailwind, long term:

If CT is indeed supplying Tesla (likely given the “US-based major EV OEM” language), there’s a free option on Tesla’s & Other Robo-taxi ramps. To make Robo-taxi economics work, Tesla needs an ultra-low-cost vehicle, hence why their new car model, is going to be simplified with less features, for pure cost optimisation.

Low cost players, who compete on price like CT, will benefit disproportionally from scaling of self driving taxi networks, regardless whether it is Tesla’s, Waymo’s etc.
I have not seen this optionality mentioned.

4. Why I’m Passing for Now

There are material concerns that prevent me from investing currently.
These fall into three categories: balance sheet quality, governance concerns, and customer concentration risk.

Balance Sheet:

The balance sheet looks clean on paper:
£66.3 million current assets versus £50.4 million total debt (1H 2025) meaning net current assets of £16.3 million.

Looking at receivables tells a different story.
FY2024, 45.9% of trade receivables were 1-90 days late, up from 26.4% in 2023. While automotive suppliers typically have higher overdue rates, this is significantly above industry norms.

This July payment (outside the 1H reporting period), has a high probability of relating to their previously disclosed largest customer — Marelli, which is going through Chapter 11 restructuring (outlined shortly, under ‘customer concentration’)

The prepayment explanation is odd, as tooling contracts typically are paid once key milestone are met, not pre-paid. This usually means:
Marelli is offering more favourable payment terms to rebuild goodwill or CT more desperately needing cash and agreeing to unfavourable payment structures. Both seem plausible.

Additional cash flow red flags:

• Significant dilution needed via 2023 capital placement to resolve cashflow struggles despite ongoing operations showing profitability.

• Jan 2025: £300,000 loan from founder to China subsidiary at 12.7% interest — a high rate for an owner-financed loan. Highlights capital need as traditional credit lines near maxed out (lender uncomfort for further borrowing) & interest rate reflective of the loan risk.

• Receivables accumulating under ‘late receivables’, whilst cash is shrinking.

Governance Red Flags

1. Mexico Ownership Structure, Buried Footnotes

Deep in the FY2024 annual report (buried across 120 pages in two brief footnotes), CT disclosed it’s CEO & COO acquired 10% of its Mexico facility for £6,000 while over £3 million has been invested in capacity expansion.

The real issue isn’t the deal structure per se, it is unclear what the intention behind this transaction is. If management buries key information in footnotes rather than clearly explaining them it raises questions.

2. CFO Turnover

Former CFO left in only 1yr 11months: Anna Brown, left after a short period, particularly compared to her previous tenure lengths (20 years KPMG & 8 years McCarthy Stone), to me this seems an exit earlier than anticipated, when considering options were set to vest from 3 years into the business.
Later updated her LinkedIn with detailed bullet points highlighting her impact, some of which to me highlight weaknesses within the accounting:

Reading between the lines, this segment is a promotion for her now consulting services whilst to me signalling:
‘I left because of the company’s issues, not my performance.’

Recent CFO, resigned with immediate effect, 28th Jan 2026. Salman Mohammed had been with the business as a Finance director, October ‘23. Later promoted to CFO before resigning within only 7months.

Both CFO exits, particularly the recent one being with immediate effect, often signals issues with reporting quality and/or governance.

3. Tooling Revenue Recognition — Key Audit Matter

Auditors flagged tooling revenue recognition as a key audit matter, noting it’s possible to be “easily manipulated.” In FY2023, tooling revenue increased 20% while production revenue declined 20%, I believe this could have been a driving factor behind FRC investigation.

Percentage-of-completion accounting for tooling contracts is based on management discretion and can easily be ‘optimistically’ estimated during difficult periods.
Particularly considering weak internal controls (per the CFO’s LinkedIn) & auditor scrutiny. Though in recent years the overall % mix of total revenue has come down.

Summary: There are a lot of red flags here, most of which frankly are potentially-used/ used to struggling cashflow & performance. This makes me uncomfortable from a going concern perspective and the reason as to why I need to see material changes to warrant investing (outlined in ‘What would change my mind’ segment)

Customer Concentration — Elevated Bankruptcy Risk

CT’s three largest customers represent 38%, 18%, and 7.6% of revenue (ref: FY2024 Annual Report). In a normal environment, this concentration wouldn’t be dealbreaking as fairly common for Tier 2 suppliers & well protected: sticky automotive supplier relationships and high switching costs for CT’s low-value parts.

But we are still at compressed volumes relative to pre-Covid.
Marelli (38% of revenue) filed Chapter 11 restructuring in June 2025.
Their is a DIP financing (loan to fund operations during restructuring) and restructuring plan suggest operational continuity (which aligns with CT management commentary) suppliers have experienced delayed payments. CT’s late receivables spike confirms this impact, given a large portion sits with Marelli.

Marelli’s restructuring is expected to resolve by mid-to-late 2026, with most lenders signed onto a debt-for-equity swap that eliminates ~$4.9 billion in debt. Post-restructuring, annual interest expense falls from ~$270 million to ~$117 million (assuming full DIP drawdown), lower interest costs, make business operations, significantly more sustainable.

However, even successful restructurings often lead to supplier losses.
CT may not recover 100% of outstanding receivables. And the process extends uncertainty through much of 2026 — how much of receivables they recover is a key metric I am monitoring.

Another client, Forvia (2nd or 3rd largest customer at 18% or 9% based on disclosed client mix) is unprofitable currently lost (€269mil) in 1H ‘25, with sizeable debt obligations.
Nissan, CT’s largest end-customer OEM (~47% exposure, as per annual report), announced major job cuts amid financial stress.

Multiple distressed customers simultaneously is not a coincidence in a market with weakened volumes; industry-wide margin compression and overcapacity.
CT’s resiliency only goes as far as its weakest partners. Even one additional customer bankruptcy could lead to their own bankruptcy.

Otus Capital Exit

Otus Capital, major shareholder which owned ~18-19% of CT as of December 2023 (bought in during capital raise), now consistently selling for 18 months.
Down to their last ~9% stake remaining & representative stepping off the board. Selling ~1-1.5% of total shares outstanding every 1-1.5 months.

At CT’s low liquidity (80-140k shares/day), Otus represents 20-35% of daily volume. This is created prolonged selling as this institution dumps their position a large part of why i’m in no rush to buy either.

I attempted to contact Otus for their exit rationale. They declined to respond. Additionally based on a brief discussion, the investor relations representative’s tone (subjective, but notable on call) suggested discomfort with the question around reason behind why Otus is selling, hinting it being ‘never simple’ whilst acting within the restrictions of what they can & can’t say.

This doesn’t prove CT as guilty of anything, but it’s telling. If the exit were purely mandate-driven (fund redemptions), a professional IR response would be straightforward, or at least without discomfort.
The reluctance to engage on both sides, suggests there may be operational or governance concerns driving the sale.

Combined with the CFO departures, this institutional exit and ~43% owners are remaining institutions locked into an illiquid £20 million microcap and with one willing to sell over 18 months at suppressed prices, suggests they’ve seen enough to want out, even at a cost.

The Core Problem:

The fundamental issue I see, is whether receivables will hold up to provide sufficient cashflow sustainably, if a healthy percentage of receivables actually convert, there will be enough to offset the majority of debt, which would significantly reduce any impacts of potential risks e.g short term sales weakness, customer struggles.

CT has already raised capital twice (2021 IPO and 2023 placement, the latter diluting by 40%). A third capital raise likely to fail, if it didn’t it would be extremely dilutive given low current valuation; institutions are unlikely to sink more funds into the business (seeing Otus exiting) and UK retail investors, have little appetite based on my experience & tendency to prefer dividend paying businesses.

If Marelli’s restructuring extends beyond expectations, if another major customer fails, or if receivables collection deteriorates further, CT would be looking at bankruptcy itself. The founder’s £300k loan at 12.7% suggests struggling cashflow & maxed out credit facilities.

For an investor focused on both:
Unreasonable, priced investment opportunities, conservative multi-baggers.
Asymmetric, opportunities, downside needs to be protected with solid fundamentals.
Currently I am not seeing solid fundamentals which can put a hard stop to capital loss

4. What Would Change My Mind:

The business has its issues. Frankly a lot are to do with the ruthlessly tough industry it operates in. I do not believe the business model is broken, CT, to me is actually one of the best ran operators from a cost & innovation perspective.

Effectively I need to see the business de-risk. Both balance sheet & customers, once the company is in a better position, any potential short term cashflow band aids to support weakness will no longer be needed. As currently used in attempt to keep the business afloat, which makes sense from the founders perspective who owns ~24% of the business. Here’s what I need to see to invest:

Note, though these points are marked separately, a lot will be correlated.
E.g market improves → Marelli financial health → better position to pay legacy receivables

1. Marelli Restructuring Successfully Completes

The debt-for-equity swap needs to finalize (expected mid-to-late 2026) with Marelli emerging confirming it will operate as normal. Whilst it is somewhat is claiming to do currently. However, CT is yet to confirm full or near-full recovery of outstanding receivables. No material write-offs for this to be proven.

Post-restructuring, Marelli should demonstrate normalized payment behaviour going forwards. I’d want to see Marelli-related receivables are current (within standard payment terms) before considering this risk resolved.

2. Receivables Quality Normalizes

Late receivables (1-90+ days overdue) need to decline from 45.9% (FY2024) to a healthier level — under 20%, in line with industry norms for automotive suppliers.

This would signal: (a) customers are financially stabilizing, (b) CT isn’t being pushed into unfavourable payment terms, and (c) working capital is converting to cash rather than being trapped in uncollectable receivables.

3. Balance Sheet Strengthens — Debt Paydown and Cash Build

Once receivables are collected, CT should use cash to pay down debt and build a cash buffer. The Mexico facility ramp should become cash-generative rather than a drag on cash by 2H 2026 per management guidance.

I’d want to see:
(a) net debt declining, ideally near zero debt, which could happen if receivables hold up
(b) cash reaching at least £5-7 million (enough to cover 2-3 months of operations)

Management has mentioned potential for share buybacks once cash flow strengthens, 2H ‘26. If they follow through, it would signal confidence in financial stability.

4. CFO Stability and Accounting Clarity

The next CFO hire needs to stick around. I’d want to see this properly addressed.

Ideally address past disclosure concerns:

A) China £300K Loan

B) 10% Mexico ownership for 10%

If management voluntarily improves disclosure quality without being forced by regulators, it would rebuild credibility for me,

5. Mexico Facility Contract Wins

The $37 million in annual contract wins needs to translate into actual cash, with cashflow improving meaningfully as heavy investment is no longer needed, for this asset which can start returning capital.

6. Valuation Still Attractive Post-De-Risking

What if the stock goes up?

I think this is a lower probability considering Otus still has around 6+ months of selling assuming a full exit occurs, however even if the business re-rates higher (within reason), I would still be looking to buy provided the above have been de-risked & still attractively rated as a whole.

Given CT’s growth profile (10%+ CAGR), margin structure (30%+ gross, 7%+ net), and competitive positioning (top 10-15% cost curve), 8-10x earnings is where peers trade at and higher if the balance sheet become healthy.

Timeline: 6-12 Months

Most of these triggers should become clearer within 6-12 months: Marelli’s restructuring timeline (mid-to-late 2026), Mexico facility ramp (2H 2026), and receivables normalization (should be evident in 2H 2025 and FY2025 annual reports), will be interesting to revisit near Q3 2026.

5. Why This Decision Reflects My Investing Framework

My core investing principle is: only invest in the very best asymmetric opportunities I find, and stay highly concentrated in them.
Which requires a level of downside protection to be concentrated. As it is crucial for me is to avoid large permanent capital losses, where I can.

Currently, when I compare CT to some of my previous investing opportunities, where I’m often able to buy at 4x-8x earnings with similar/higher growth but no large customer bankruptcy risk, stable CFO, clean governance, and strong balance sheet.

I would simply rather have been more concentrated in such positions.
As always any questions you have, I look forward answering
Thanks,

I read all them from, December 2023 — July 2025 (~19months), which gets us to the 1,000+ number.

Two stood out in a league of their own: Sezzle (SEZL) and NextNav (NN).

Both became multi-baggers. Both were completely different businesses. But the write-ups shared something critical that separated them from everything else.

Links to the full VIC write-ups are at the bottom.

WHAT MADE THESE STAND OUT

The authors knew the history—not just the numbers.

They understood:

1. The founder’s track record across previous ventures—not just current execution

2. Intricate industry dynamics, that most surface level screening would miss

3. Failed deals and structural events & history which created the mispricing — then able to decipher between Rational/Irrational

Sezzle: The founder story

The write-up detailed how a VC investor (Lee Brading) had backed founder Charlie Youakim in his previous venture (Passport), then followed him into the public markets as an investor in Sezzle—despite being a private equity investor—purely because of the individual running the business.

They documented his track record: mechanical engineering degree, software dev in payments, MBA, co-founding Passport (2010 exit in 2015), the pivot when Sezzle's original idea failed.

Then the author explained the BNPL credit market dynamics from first principles—how consumer credit evolved, what caused the 2022 collapse in BNPL valuations, why Sezzle's turnaround to profitability was structural (cut $113M in losses in 2 years while maintaining growth), and what the unit economics looked like in the value creation phase as they shifted from land-grab to significant profitable growth whilst the company was still priced at near bankruptcy.

NextNav: The asset play disguised as a moonshot

Majority saw: SPAC burning cash on flying car technology.

The write-up revealed it was actually an asset play: 8 MHz of clean, contiguous low-band spectrum covering 93% of the US population.

The author broke down:

1. Comparable spectrum transactions (T-Mobile/Comcast at $2.21/MHz-pop, Anterix precedent)

2. Quality differentials (NextNav's spectrum was "plain vanilla" with no incumbents vs. Anterix's county-by-county slog)

3. Government incentive structures (FCC wanted more broadband spectrum, had already repurposed similar 900 MHz band)

4. Why key management would enable the asset sale: new CEO sat on NTIA Commerce Spectrum Management Advisory Committee and FCC Technological Advisory Council

They identified Neil Subin (Lloyd Miller III's partner, Straightpath multi-bagger) who owned 7.5% and documented his track record across dozens of spectrum plays—backed by the best capital allocator in this niche.

WHY THE OPPORTUNITY EXISTED

Sezzle: Exchange transition forced selling. Uplisted ASX to NASDAQ, but a retail dominated shareholder base of 10,000+ Australians lost interest.
This was no longer "their" company.

NextNav: SPAC stigma + $35M annual cash burn + “flying cars” in the investor deck = immediate pass for 99% of investors.
Which usually is the right move, though not in this instance where you have a hidden asset worth $1Bn+

THE PATTERN

Both had strikingly obvious asymmetry once you looked past the surface:

1. Downside protection was structural:
Sezzle had just turned profitable after cutting $113M in losses, was trading near tangible book value, and had clear buyout interest at higher than, then current valuations (Zip's failed $330M acquisition). NextNav's spectrum was worth $8-18/share in conservative scenarios vs. $3.76 stock price—even accounting for $35M annual cash burn.

2. Multiple paths to 5-7x upside without heroics:
Sezzle: Multiple expansion from 14x to peer average (20-25x) + 20% revenue growth + already occurring significant operating leverage = 3-5x
NextNav: Spectrum sale at $1.30/MHz-pop (conservative) = $13/share. At aggressive $2.21/MHz-pop = $23/share. In comparison to $3.76/share both scenarios visible multi-baggers without bull-case assumptions.

3. Structural reasons for mispricing:
Forced selling (Sezzle's ASX retail exit), SPAC stigma (NextNav's surface-level red flags), investor type mismatch (Australian retail vs. US microcap, SPAC tourists vs. spectrum experts)

WHAT I LEARNED

Conservative multi-bagger opportunities require knowing vastly more than surface-level analysis suggests.

These authors didn't just know the numbers—they could explain:

• BNPL history from Great Depression layaway programs to modern Pay-in-4

• Founder's family (5th of six immigrant children), engineering degree, previous exit

• Competitor economics (Klarna undercutting on fees but burning cash, Affirm shifting to interest-bearing)

• Exact approval boards (NextNav CEO on NTIA Commerce Spectrum Committee and FCC Tech Advisory Council)

• Investor track records across decades (Subin's 10+ spectrum plays since 1990s)

This depth eliminates blind spots. When you know more than industry insiders, the risk of missing thesis-breaking information drops dramatically. The multi-bagger math works because you've stress-tested every angle.

That's the standard worth aiming for.

---

VIC Writeup Links:
Sezzle - Author: templargin

NextNav - Author: MJS27

With a market cap of $311 million, TTM Earnings of $51.7million, earnings expected to exceed revenue growth, all whilst paying you 7.1% Dividend yield, this business is too cheap too ignore.

Index:

1. Introduction

2. Acquisition Rationale

3. Loan Breakdown

4. Deposits

5. Business Services

6. Hidden Cost Advantage

7. Financials

8. Management

9. Shareholders and Liquidity

10. Growth Runway

11. Risk & Downside

12. Valuation & Why Book value doesn’t work

    Subscribe for more Deep Dives

1.Introduction:

NewtekOne Inc, founded in 1998 by current CEO, Barry Sloane, owning c.5% of the business, the company began by serving American SMEs through technology solutions such as: Payroll, Accounting, Merchant processing / payments, Web hosting & cloud, Insurance, etc.

Whilst the business did offer loans (which is the current primary offering) such as the SBA 7a loan program during this period, this hadn’t yet became their main focus, it was only up until the business solutions segment peaked around the 2009 period and after a few years of stagnation/slow decline due to increased competition, the business successfully pivoted in late 2014 into primarily offering loans for small/medium sized businesses with the prior, still profitable services segment continuing & greatly aiding with retention and cross selling.

At this point the business opted to become a regulated investment structure, a Business Development Company (BDC), this structure required to return 90%+ of profits back to shareholder & any new capital raised must be at least 45% funded by equity. Newtek had done fairly well considering the majority of its capital was funding annual dividends of 6 to 8%, on a dilution adjusted basis and the stock managing to over double during this period.

To highlight the amount of capital the business had returned, in the last 4 years, Newtek paid out $228million (market cap at today’s writing is $326million), though not a vital valuation metric, for me as a reader would remove doubts of potential fraud/too good to be true.

In 2023 Newtek acquired National bank of NYC (now NewtekOne) with deposits of ~$137 million and transformed Newtek into a Bank & Hold company, dual structure, this move alienated their shareholder base at the time who were investing purely for the high & consistent dividends, would then be forced to exit the position (no longer meeting their dividend targets) as the pay-out ratio was no longer going to remain anywhere near 90%+, especially due to the business now being a bank, especially as management made it clear that regulatory approval was required, for any dividends to begin with.

This led to what was a painful process as many investors were frankly pissed off (reading online forums) at management who had a good thing going so why fix it? Which I will address later. However over the next few years as the shareholder base & analyst coverage completely rotated, for this non-fundamental related reason, the stock declined from the peak of $38 to $9.50 at the most recent low in May 2025 — Hence this opportunity presenting itself

2.Acquisition Rationale:

As mentioned earlier, Newtek had acquired National Bank of NY City in Jan ‘23, though the initial agreement started in August ‘21, for $20mil price with delay due to regulatory approval. This brought on $137million, in mostly local deposits. Here we will go through:

1. Why Investors didn’t like the deal?

2. What was the rationale behind the acquisition?

3. How has the deal worked out, 3 years later?

Why Investors didn’t like the deal?

To put this simply, Why Fix something which isn’t broken? But to really understand we again, go back to the history of the business having been a BDC (Business Development Corporation), which is an investment structure that required 90% of earnings to be paid out as dividends, meaning the types of investors who had been investing into the business were there for a tax efficient high yield vehicle to which NEWT had consistently generated 6%-8% dividends annually (net of dilution) - Pretty good from a pure yield investment, perspective.

This all changed as the business had become classified as a bank meaning, firstly that any dividends would need approval and even if they were approved it certainly wouldn’t be 90% payout rate - for previous investors it meant uncertainty of whether there was going to be a dividend and if there was the yield would significantly reduce. Albeit shortly after the acquisition Barry Sloane did mention in an earnings call that a proposal was set forward to regulators to approve a 50% Payout rate. Regardless the focus had changed from yield first to capital appreciation, by way of growing earnings & Book value; naturally this would mean a rotation of the shareholder base.

Additionally this presented increased complexity as the business is now split into a dual structure where:

A) Bank segment of the business which takes on customer deposits and originates loans, this segment uses CECL accounting, as per bank requirements.

B) The Holding company, holds a fair bit of it’s legacy loans which are into runoff, but also partakes in the packaging & selling of its new loans, as well as the business services segment, mentioned in the intro. Hold Co, remains using the original fair value accounting method.

It’s worth saying that most loans had also been transferred into the new banking segment, due to the different accounting styles, have made the numbers more complicated as there is no longer a fair, year to year comparison over the short/mid term.

What was the rationale behind the acquisition?

The business had previously operated as a private credit lender essentially borrowing institutions capital (predominantly from Deutsche bank & CapitalOne) as well as from bonds & other debt structures offered to investors as well as through dilution, the business was profitable, whilst being in a tax efficient structure however there were 2 main benefits from this change:

A) Previous cost of capital was generally between 6.5% to 9%, depending on interest rates (now at 9%).

This was still good considering the business could make a Net Interest Margin (NIM) on originated loans of ~2% however with the business being able to take on deposits, if this could get to around 80%+ of funding, like current peer LiveOak, the business could have a cost of capital more in the range of 3% to 4% meaning NIM could easily expand, potentially even to 6%+. Whilst this sounds crazy, its doable another peer (though in the British market so they don’t compete), is DFCH, another example of a Private Credit Lender turned Asset Lending Bank model has NIMs of 7%+ using this same playbook.

This obviously provides a significant cushion should rates rise sharply as well as to have significant increase on operating margin, if NIM doubles I would be surprised to not see net margins at least follow.

B) The BDC structure required for a minimum of 45% of all new capital invested to come from equity

A significant weakness of the BDC structure which inhibited the business from growing in any major way, as earnings couldn’t be retained (90%+ payout minimum on earnings), leaving the only real way to access to meet the 45% equity quota, to reinvest into growth - was through consistent dilution, which would obviously hinder the businesses ability to grow in any real way. Said another is below in Barry’s (CEO, founder) words, on an earnings call:

“At the BDC, only 55% of that loan gets funded by our warehouse line. At today’s cost would be almost 8.5%, Maybe 8.75 percent. So and 45% has to be funded by selling shares, by selling equity, By diluting adding more share count to EPS. So even though you got to pay tax in a bank (structure), the math just doesn’t even compare. That’s why one of the reasons why we did the transaction.”

By having changed to a bank structure meant the business could no longer act as a BDC, therefore allowing the business to fund grow without needing to constantly dilute.

Additionally it is worth noting from the business segments Newtek has opportunities to significantly cross sell as it would go on to auto-open accounts free of charge. The business was also able to offer high rates for depositors due to the significant discount in comparison to interest paid on capital from private credit lenders.

How has the deal worked out, 3 years later?

Initial deposit base of $137mil to now $1.04bil! Deposits outpacing loan origination (growing 17% YOY) whereas deposits still growing at 30% YOY. This is allowing the business to access significantly cheaper pools of capital. Meaning that growing deposits now make up 53% of funding, NIM has also grown from 2.1% to 3.1% in the last 2 years.

You may be thinking shouldn’t the NIM be closer to 6% you mentioned earlier? In short, normally Yes and I will go into the reasoning in the Loan Breakdown segment.

Cross selling has worked, and has significantly contributed to the roughly 7X in deposits particularly as a lot of the services require cash to be parked somewhere, so why not conveniently save it in the Newtek ecosystem where, as a business, we will need to do our payment processing, payroll, taxes, insurance etc from here anyways.

Newtek has been cautious of deposits being funded by large institutions and seeing them more as a ‘liability’ is my assessment based on earnings commentary, particularly with Silicon Valley Bankruptcy being fresh news at the time of scaling deposits. The business has been conscious of large institutions who are mobile with their capital, seeking & moving capital to the areas with the best yields —And rightly so.

Therefore by avoiding this risk; whilst factoring in the Stickiness of their services ecosystem; and paying top of market rates (organically attracting smaller depositors who find Newtek because of high rates and no fees on aggregators) has de-risked Newtek’s deposits.

Paying higher rates isn’t an issue as Newtek has the economics to afford paying depositors well and don’t face the risk of their capital flight in the same sense, their checking account only makes up for about 20% of deposits too, overall their capital similar to their loan strategy, depositors are well diversified.

3. Loans Breakdown

Brief overview on loan types:

SBA 7a: Small Business Administration loan generally under $5mil with 60-90% government backed (~75% on avg), used for working capital, equipment, business acquisitions, or real estate. Terms up to 25 years for real estate and 10 years for equipment.

SBA 504:Owner occupied mostly real estate loans mostly up to $5.5mil, 40% government backing, with usually 10% deposit.

ALP (Alternative Loan Program): Non-traditional lending, funded on the based of collateralisation and personal guarantees on the loans protecting Newtek. Where rare defaults have occurred, recovery rate has been very high. As a result to-date, Zero losses in this segment have been realised, reflecting both the rigorous collateral coverage requirements and high quality of underlying assets securing the loans, having spoke to management who confirmed this, therefore would see ALP to be self-sufficient without Government backing.

CRE & C&I: Traditional commercial loans, generally lower yielding, (Commercial Real Estate loans and Commercial & Industrial loans).

CRE finances investment properties based on rental income with a typical deposit of around 25%. Though a significantly smaller segment C&I provides working capital, equipment financing based on cash flow. Loan terms also vary based on the asset being loaned against.

Newtek’s Loan Underwriting Approach:

• Little /No Covenants - there is high amounts of pushback from business owners who want to freely use capital. Hence by not imposing such rules, is a selling point which allows Newtek to achieve on the below terms:

• Very high collateral requirements, first seniority on assets (rarely secondary) in terms on lien rights priority & personal guarantees from business owners with a 20%+ equity ownership of the business.

• Yes, the business loans have higher charge offs (compared to traditional loans which is to be expected for the type of loan) however when accounting for the increased interest charged, which Newtek then uses to build up Allowance for Credit Losses (most recently about 8%+), it still makes sense to do on this basis.

• Over last couple years, alongside peers, switched to offering primarily, variable interest loans, for example on SBA 7a, currently Newtek looks to charge Prime rate +3%, this is in particular after lenders in this segments were burned by offering mostly fixed interest loans in ‘22, locking in a low rate. Which then ended up having negative NIMs as Interest rates increased sharply. This situation would be prevented if it were to occur again. I estimate this to be around ~75% (for SBA 7a), though there isn’t any hard data backed up, this can be seen by which loans are primarily advertised & through commentary in an article with Newity (another SBA lender) “Though some lenders offer these loans with a fixed interest rate, it is uncommon.”. As well as confirmation that Newtek hedges on its fixed loans.

High ACLs, High collateral (first liens), Personal guarantees, Government Guarantees & Variable Rate loans has been the combination which allows for Newtek to receive High Yield Rates at Lowered Loan Risk in their portfolio.

Newtek is able to achieve these terms in exchange for having more flexibility offered, to business owners.

• SBA 7(a) loans, Roughly $407 million on the book, with $1billion in originations estimated for 2025 YE. Of which majority will be sold off to investors.

• Commercial Real Estate (“CRE”) , $231 million. Commercial & Industrial (“C&I”), $51 million: totaling $281million whereas by year end the portfolio is estimated to be $464million by 2025 YE — Not a priority as this is lower yielding segment.

• SBA 504 loans & Alternative Loan Program (“ALP”) loans: No clear breakdown on exactly how much is held, SBA 504 is primarily sold upon origination. ALP program is primarily bucketed under SBA segment, with $500mil estimated to be originated for ALP. Whilst SBA 504 is targeted to have $250mil origination.

The business is constantly originating loans in significant capacity, beyond a bank of their loan book scale (mainly because they’ve not had chance to grow their book-value properly for the last decade under the BDC structure), then selling excess loans by packaging into a portfolio of loans to investors as this has the highest return on capital, that the business has available, I see this part of the business as a gem considering, not much of the loan risk is held. Newtek is able to sell a portfolio of loans at often an 8%+ mark up, with the ability to repeat this model 3x up to 4x per year on the same capital giving you ROCE of 20% to 30% on this segment, almost unheard of, for banks.
Additionally as the business continues to scale its origination volumes, Newtek will be able to package more loans into a package, meaning loans will be more diversified for investors, increasing demand for this offering, allowing a quicker sale turnaround.

For investors, who looking to work out exactly what their book is made up of, this is a wildly moving target though still of relevance — it is not as important in comparison to traditional banks.

Newteks’ gain on sale has been a large consistent reoccurring transaction for the last decade. Therefore I have included the origination volumes (image above) provided for each segment, as loan origination and loan underwriting approach provides a clearer image of Newtek’s loan operations whilst still providing an idea of the loans held. Numbers have been provided directly from Bryce Rowe (IR at Newtek), Origination Estimates Source

This shows that 51% of the loan originations are from SBA 7a, ALP is 25% and SBA 504 with CRE and C&I make up the remaining c.24% approximately.

Shift in Lending model:

As mentioned earlier, Newtek switched from a mainly fixed loan offering to a variable offering on its largest segment in particular, SBA 7A, this is due to the business alongside the majority of its competitors being caught offside, as interest rates rose significantly and sharply in 2022 & 2023. Meaning that the loans at the time which they would have wrote at 8% interest rates looked like a good deal when their cost of capital was at 6%, however only a couple years later did the cost of capital for Newtek on these same loans increase to 8.5%+, meaning even without any loan losses the business was losing money on the loans (Near Zero/Negative NIMs). Note: the only competitor which hadn’t lost money on this segment during this period, pretty much sold all their loans upon origination — though receiving less earnings per loan in the process.

This had happened to competitors too, as a result this market more or less shifted to offering variable loans as the primary offering meaning loan structures today look something like this: Prime rate (Currently around 8 or 9%) + 3% spread on the loan, the prime rate would then adjust along with the rate set by the federal reserve. Additionally I estimate that in the SBA 7A space, no more than 30% of loans are sold on a fixed basis, of which are hedged.

During peak losses for this segment, the business had been generating $70mil per year in net income whilst losing $10 to $15mil on these loans they’d wrote per quarter, so significant losses, this also explains why NIMs had only grown from 2% to 3% in the last 3 years — Profitability increases masked by short-term problem.

Further complexity:

Not only are the majority of loans sold, or waiting to be sold. The portion of loans which are HFI (Held for Investment) are split between the original private lending structure of NSBF(Newtek Small Business Finance) which has had most loans transferred into the new book (associated with the bank) and the remaining having gone into runoff (in the original structure). Loans which had been transferred are stored under the banking structure, and both have different accounting methods used on the loans.

NSBF (original structure) uses a fair value assessment, meaning it looks at the most recent comparable sale and then adjusts the price of Newtek’s existing loan book to be inline with that sale price. Whereas CECL which is required by deposit taking banks, sets aside a loss reserve on the book, then writes it off - usually making the loans seem less profitable at the start of their life as more of the reserve gets built upfront.

Selling Vs Retaining loans:

Usually the business, sells all the government guaranteed portions (so about 75% of each loan) to investors currently priced at about a, 6.5% yield, valuation to investors on these almost ‘risk-free’ assets — due to Gov backing.

Newtek would historically retain 50% of the uninsured c.25% portion — as in the part which gets hit first if any losses occur. This would be split 50/50 in a joint venture, but in the most recent quarter the business decided to retain the whole amount and likely to continue. Reason for switching to wholly owning the risk in a recent example is that the loans are risky which is seen as unattractive but management feels is priced at too high of a yield to pass up, even when considering risk. An example was provided in the Q2 ‘25 earnings call where the ALP residual/uninsured, of the secured loan book was fully held by Newtek.

The portfolio had been ascribed a 20% loss severity with a 15% probability, giving a 3% expected loss (20% x 15% = 3%). The loan book was priced at a 14% yield already factoring in a 3% loss reserve. In the assigned loss scenario, if losses actually hit 20% across the entire portfolio instead of the expected 3%, the investor would only lose about 3% total (20% loss - 3% Buffer - 14% expected yield = 3% actual loss), not the full 20%. Though uncertain, the risk had been priced in to make the loan low risk in this sense, though still uncertain & ALP having very strong collateral.

In general Newtek has always taken the approach of being conservative with its provisions for credit losses, and yet the business still shows immense profitability.

Provisions for Credit losses (for the loans held under the banking structure), being as high as 8.3% in a recent quarter, although net charge offs’ have rose to 3.1% this is normal considering that these business loans are generally riskier, though with higher yields which price in this risk. And that the loan book had been ‘rebuilt’ for their new Newtek Bank structure, it is expected for the loan book to ‘season’ over time in the CFOs own words — meaning any losses before Q4 2024, are unrealistic, given not enough time for loans to start defaulting.

Government Guarantee Risk

The guarantee itself is a bet on the pillars of America itself, to fulfil on its obligations. In terms of the risk of Government backing disappearing/significantly reducing which would adversely impact likes of SBA 7a, I see this to be a highly improbable for the following:

• SBA program, is seen as more of a gateway into private and public partnership, I do not see this closing, particularly from a capitalist perspective.

• Bipartisan support on the SBA program, it is seen as one of the limited ways to support small businesses, an indirect way to increase employment.

• Over the 70 years, this program has been through, all sorts of administrations with no issues though not a guarantee shows durability, nor has there been mention of changing the program.

In summary, this is a point of reliance for Newtek and will always be a risk, though I still see the model working even without guarantees, it simply means cost of capital will become higher. The imbalance, unsatisfying the demand for small business lending would become even greater if there were to be significant changes to SBA guarantees, on the off chance. Even so the business would be primed to pivot to ALP loans which aren’t reliant & have proved success alongside more of the traditional CRE and C&I loans, though these are lower yielding.

4. Deposits

As mentioned earlier on in the acquisition segment, since acquiring Bank of NY deposits, started with $137mil deposits, since Jan ‘23, and today at over $1bil, more than a 7x in deposits!

Segment is continuing to grow at 30% YOY, at controlled pace as management is conscious as certain deposits are more harmful than beneficial, being particularly cautious to exclude taking on deposits from large institution who are becoming increasingly mobile when it comes to chasing yield on their capital.

Capital is well diversified amongst smaller deposits, which are more sticky, with 70%+ of total deposits being within the $250k FDIC insured amount for bank deposits — providing higher security as most depositors wouldn’t be as likely to trigger a bank run, unlike with Silicon Valley Bank. Additionally management is satisfied with the percentage of deposits stored in the low yield checking accounts, being about 20%, whilst this has a much lower yield payable so more profitable, the capital is less stable. Personally I am satisfied with the mix, considering that 70%+ comes from High yield and CD accounts. The high yield accounts, which Newtek already pays the upper range within the market makes this product less vulnerable to migration, away to significantly more attractive yields. Management is not blind to any ideas of loyalty for their brand to afford to pay low rates.

A significant opportunity remains for deposits continuing to become a larger amount of capital funding the business, particularly as the pace of deposits growth at 30% is outpacing revenue growth (which I’m using as a proxy for loan growth) at 17%. Currently at 53% of capital used is funded through deposits, wouldn’t be surprised to see this number reach 80% over the next 4 years approximately, assuming a similar pace of growth.

The reasoning behind the assumption of a similar pace of growth is: due to growth mostly being organic by high yields being shown on aggregators with the Newtek site then setup to convert new depositors. Additionally seeing that Newtek at times has intentionally opted to control the pace of certain deposit products it offers - to me is enough to show there is sufficient runway ahead.

Though there are not many players who have incorporated deposits into the type of loans such as SBA 7a. Those who have (such as LiveOak), generally worked even at higher deposit percentages around ~90% of capital funded through deposits. The government guarantee on a large portion of loans, adds stability which sets the bank up nicely to hold deposits, see no immediate risks such as regulatory preventing the business from growing deposits as a larger percentage, considering peers have done this.

Non-Deposit Funding
Currently the remaining 47% of non deposit capital is funded through partnerships with Deutsche Bank, CapitalOne, Goldman Sachs etc. credit facilities at c.9% cost of capital and capital is also raised through non-dilutive equity offering at a 8.5% yield and other previous raises.

The rates here are high comparatively, however more or less inline with traditional private credit cost of capital rates, this only highlights how significant deposits are for Newtek profitability.

5. Business Services

As mentioned, earlier Newtek has a business services segment consisting of: Lending, Payroll, Insurance, Technology, Payments, Accounting, Merchant Processing, Web Hosting & Cloud.

This segment had initially been the core business, later having pivoted in 2014, (as outlined in the intro) however to date this business segment acts as an ecosystem for businesses by providing general business services all in one convenient place. This segment generates $12.4mil in EBIT per year, which is likely to be worth a large portion of the market cap to date of $300mil+ as a standalone business, if you consider a p/e of ~20x is pretty reasonable for a growing, high margin tech business. For Newtek this is a stable segment which adds some diversification and most importantly cross selling for lending & deposits.

Cross-Selling:

Through Newtek’s services ecosystem and more particularly through it’s; payments processing as a subsegment — which processes over $5billion in transaction volume annually. Newtek has been able to take advantage of existing customers to auto-open free accounts to cross-sell deposit accounts, which is then advertised on their website through a Zero-Fee calculator, to show how much the depositor would save, by not having to pay for account fees, wire transfer fees etc, as it is a true no-fee account.

6. Hidden Cost Advantage

I’ve purposely been cautious avoid the term MOAT as this is a high bar to cross. Newtek’s advantage lies in that it is a:

Pure Private Credit SBA Lender turned Bank funded through low cost deposits competing against SBA lenders having to rely on comparatively expensive institutional capital.

To give an idea in the current environment pure private credit lenders are paying 8.5%+ on the cost to borrow capital which is still ok, given business loans can often be 12%+ whereas LiveOak (a peer who, Newtek is on-track to have as much in terms of deposit funding) who provides the same loans has a cost of capital of ~3.83%. Though this model isn’t completely unique to Newtek. But it is important enough to separately highlight in this segment.

I agree with Barry (founder) who alluded to saying: SBA lenders who are able to use deposit capital are unique to the remaining 95% of competitors.

Could this be replaceable, in theory Yes, in practicality, not many could succeed or would want to make this change, there are three main types of competitors which could potentially clone this model:

A) SBA Private Credit Lenders - They would require decades of experience to prove to regulators that the business is capable of sufficiently managing credit risk before being able to even consider taking on deposits.

To acquire/receive a banking license itself, would take about 18 months in regulatory approval, the pain and risk which comes along with investing time and capital into offering deposits is something, which most shareholders are reluctant to put up with particularly as private credit businesses are already doing pretty well, and do even better during recessions, so there wouldn’t be a situation where change is needed out of necessity either - why fix something which isn’t broken?

Should the transition workout poorly, managements own jobs could be at stake, this comes back to their personal interests which could hold them back.

B) Regional Banks - Whilst there is often some SBA lending provided, this is usually more so on an opportunistic basis, regional banks wouldn’t be able to lend a significant portion of its book, due to regional banks being licensed under, a General-purpose depository institution, which wouldn’t have authority to be concentrated unlike the way Newtek is.

C) Large banks - The scale of SBA loans isn’t large enough to attract financial giants, it’s seen as ‘too much’ work for the scale of the loans and the TAM. The loans are seen as ‘risky’, due to higher defaults - this would also make it unlikely to get regulatory approval as the majority of these bankers have no/not enough of a track record.

In fact, large banks and financial players have been a significant referral source for Newtek over time, this has been a very profitable relationship for both sides, so there isn’t much incentive to again, fix something which is working well.

It is not impossible to become an asset backed lender taking on deposits, however it is not an easy position to get to. This barriers are enough to filter out 95% of competitors.

7. Financials

Here we are going over Newtek’s financial performance, Liquidity, Revenue, Evolution of Net Margins and where I see margins going over the next few years.

Revenue Breakdown and Growth: Business has grown fairly stable manner, growing revenue at ~31% CAGR over the last 5 Yrs.

Newtek as a counter-cyclical business; benefits immensely from recessions & economic weakness, due to traditional banking players, during these periods tightening up their lending appetite leading to private credit and similar players being able to take market share, this has clearly been shown during FY 2020 & 2021, during the peak of covid uncertainty. Even outside these periods where Newtek benefits from explosive growth, Newtek had been able to CAGR revenue at 20% during ‘more normal’ periods, with recent slowdown leading the business to a growing revenue at a ~17% growth rate for the foreseeable.

Net Income: Margins have compressed over time, with N.I margins start at 36.5% in 2020 in comparison to TTM 2025 being 14.7%, though Net income has still CAGR at ~9% over the last few years.

Though net margins in 2020 started at 36.5%, this is higher than their normalised net margins when considering the business has had substantial loan growth in 2020, meaning a large portion of the loans underwritten hadn’t yet started having any defaults, as a result temporarily inflating earnings.

Whereas margins fall sharply to 24.6% in 2022 and 17% the following year in 2023, this is due to:

1. NIM (Net Interest Margin) being squeezed, this was due to the majority of loans at the time being originated on a fixed ~7%+ interest rate (at the time), which was ok, until Federal Reserve started sharply raising interest rates ~4% over these couple years, meaning a large portion of these loans became unprofitable. For context Newtek’s non-deposit cost of capital today is at around ~9%.

2. Deposit segment, took up attention & capital investment to operate & scale, Newtek had been less efficient during period.

3. Deposits grew rapidly over this period, at a much cheaper cost of capital ~3.7%, which is in part the reasoning the business has remained fairly profitable, even as a large portion of loans had become unprofitable.

Their margins over the last couple years have stayed at the 15% range, competitors in this space have historically had 15% — 25% net margins.

For me, Newtek has a clear line of sight of having net margins stabilise at a minimum of 25% if not higher over the next few years as:

1. Loan losses continue decreasing having gone from a ($14mil) loss per quarter at the peak to ($4mil) loss per quarter, in the most recent quarter (Q3 2025), on these same loans, originated in 2022 and 2023) — Ref image below.

2. No more drag from setting up deposits & cost associated with restructuring post acquisition.

3. Deposits continue increasing as a percentage of capital funding, currently funding 53% of capital and I’m projecting 80% by 2028 YE (continuing ~30% CAGR)

4. ~70% of new loans originated are on a variable rate basis (significantly less impacted by rate increases) and even for the remaining 30% which are originated on a fixed basis — we have just come off, sharp rate hikes and more likely than not, rates will continue falling, so very little risk of NIMs being squeezed for the foreseeable.

Capital Liquidity: Assessing the financials, we are able to see Newtek holds, $381mil in Cash&Cash Equivalents, with ~$450mil in loans held for sale (after adjusted for allowance credit losses), along with a visible $360million credit facility in loans used for funding loans such as ALP, for the purposes of being packaged and sold to investors. Credit facility amounts are split between Deutsche Bank, CapitalOne, Goldman Sachs which are $170mil, $100mil, $90mil respectively alongside other smaller amounts through other players where the exact amount hasn’t been disclosed. Though this has only been setup to fund loans to then sell on to investors, it reduces the capital-intensity burden, putting Newtek in a stronger cashflow position.

The point I’m illustrating is that Newtek has sufficient capital liquidity to de-risk, from any potential short term liquidity issues which could occur, from their (though stable) deposit base of $1,060mil, alongside ~$60mil further in other current liabilities.

Regarding the credit facilities Newtek however a large majority are to be paid upon loan portfolios being sold and would be paid from the sale capital, this is observed in the cashflow statement under ‘Proceeds from sale of loans held for sale’

Meaning Newtek has actively positioned itself to having a more stable deposit base and even in the instance should a large amount of depositors leave overnight, I’m comfortable with the liquidity position the business is in.

8.Management

Here is my management commentary, whilst I have spoken to management previously, keep in mind this is still from an outsider’s perspective and it’s subjective, but I’ve provided my rationale below. Ultimately, readers must form their own judgment.

Overview:

As shown in the table above, management are qualified, in the sense each individual has decades experience working in respective and relevant fields to that of their current roles. this also includes independent Directors who sit on the board, they have easily 30+ years experience.

Ownership:

Barry Sloane, founder owns ~4.7% of the business, not a small amount when you consider how much dilution has occurred (with previous BDC structure) and the limitations of large ownership stakes in a bank. This comes across in earnings calls, Barry makes it clear, he is aligned interests of shareholders.

The rest of the management team, do not nearly own as much. I do not expect loads to be owned, as they are non-founders however their ownership stake is relatively low at about 1 to 2x their salary, though nothing to sneeze at, in an ideal world would like to see this higher, as interests would be more aligned. In fairness among the execs, some new hires have been made in relation to the new Deposit Banking segment, so haven’t necessarily had time to build ownership.

Weaknesses:

1. Incentives are weakly aligned as the majority of directors only have 1 to 2x salary in stock.

2. Based on research including the likes of Glassdoor, employee satisfaction is quite low compared to peers, in particular issues stemming from: controlling/monitoring systems (though appreciate difficult to resolve considering primarily WFH culture.)

   Alongside Software’s & Pay issues (below market rates) - could imply less operating leverage ‘efficiencies’ than management has previously suggested on a go-ahead basis as the business scales.

3. Key-man risk: Barry (64) is still the central figure on earnings calls (though has shifted somewhat shifted more recently) and, based on Glassdoor commentary, shapes much of the culture. That level of concentration makes the organisation feel less robust than teams with multiple driving leaders, even though I’m told the day-to-day operations are not dependent on him.

   This doesn’t imply a “one-man rule.” Early in COVID, the CFO pushed for full remote-work readiness despite Barry saying he didn’t expect large-scale shutdowns. The team moved forward regardless, which ultimately proved correct.

4. Management targets & Barry’s commentary on earnings calls can at times be a bit too optimistic, do not see this as an issue — for me this important to be aware of as an investor. Especially as Barry is a skilled salesman, which is unsurprising considering he started selling ceramics on the street at 11 years old.

   E.g. A 2023 earnings call setting a benchmark ROTCE of 30%, in reality the business actually achieved 24% — optimistic in hindsight & did get revised later on (however still a phenomenal result!)

Strengths:

1. Savvy capital allocators whether it is using non-dilutive equity issuance to raise capital at slightly cheaper rates (whilst being favourable in terms of CECL liquidity requirement) alongside other options such as institutional facility. Great track record when making a rare acquisitions, e.g loan segment which is now the core business segment had only been a fraction of revenues at the time of acquisition and was critical in allowing Newtek to pivot in 2014. The more recent Bank acquisition where Newtek grew deposits 7x+ in 2 years. However their acumen can even be seen in decisions to deciding to split the loan book favouring higher yielding loans whilst using high collateral to mitigate risk.

2. Experienced management all having 20+ years experience across the board (image breakdown above)

3. Prudent from a risk perspective such as: high loan loss reserves, loans being well diversified & conservative loan underwriting by utilising personal guarantees and collateral with healthy LTVs. Again can be seen with deposits Newtek; actively limits the number of significant corporate deposits which are often very mobile and instable acting more as a liability than an asset

4. Strong historic track record of executing, the business has managed to pivot continuously over time and sustainably grow, but even when it comes to quarterly earnings almost always hitting target, with occasional beats on earnings. Integration of deposits and actually successfully executing cross sell opportunities large part of the reasoning behind deposits growing 7x in 2 years.

5. Proven ability for management to act in long term shareholder interests, this is highlighted in the ‘sustained cost advantage’ segment, talking about the large number of hurdles when it comes to changing to a bank depository structure. This is also the reason as to why I’m not too concerned with share ownership, as the only concern for it in the first place is that it often helps align to long term shareholder incentives which are already being actioned on.

My honest take: Not necessarily an Steve Jobs at Apple Era (nor does it need to be), management has consistently executed to a high-level. Whilst they aren’t perfect, strong track record speaks for itself.
Barry is definitely aligned and has confirmed, as per our call that, he loves his role, with no intention of stepping time anytime soon.

9.Shareholders & Liquidity

Management ownership across the board is limited at approximately 7.1% of which founder Barry Sloane makes up 4.56% after purchasing $300k+ worth of stock in the last few months. Remaining Executives owning smaller amounts about 1x to 2x Salary in value.

Institutions own 46.7% of total ownership, with the top 10 institutions accounting for 31.74%.

The largest shareholders are: Patriot Financial Group, BlackRock, Vanguard at 8.06%, 7.74% & 5.44% respectively, both mostly purchasing through their passive ETFs, Barry Sloane, 4.56% ownership, as mentioned previously.

Overall ownership structure here is healthy; with no controlling shareholder risk of self-dealing is significantly reduced, as well as no large owners with a greater than 10% stake, which can often cause forced selling overhang unlike BlackRock & Vanguard who are passive, reducing this risk. Whilst Patriot Financial Group is a PE player known for specialising in the banking industry and has active participation in providing their expertise, aiding Newtek long term.

Stock Structure: (Total 26.32million shares outstanding)

• Common Stock, 1 vote per share, 26.3 million shares outstanding, this is pretty much the entire stock float.

• Series A Convertible Preferred — Previously: 20,000 shares, issued at $1,000 each under the Series A offering, injecting $20mil in capital. In return for an annual yield being 8% or matching common stock dividend whichever is greater. This has been sacrificed in exchange for 8.06% ownership for an additional $10mil capital injection.

  Total $30mil for 8.06% implying a $372mil valuation / ~$14 stock price buy-in.

  Whilst this is capital isn’t necessarily required the support from Patriot is valuable, shareholders aren’t losing out as expertise is brought in exchange. Such as managing partner of Newtek banking being Kirk Wycoff from Patriot.

• Series B Preferred, depository shares, are non-dilutive (non-convertible), used to raise capital with 2,000,000 shares (each 1/40th of a Series B Preferred share), at 8.5% yield (then post 2030 resetting to 5 Yr Treasury + 4.735%), this structure was chosen over other forms of capital raising such as debentures, reason being is, this helps with capital strength requirements as this type of issuance is classed as equity instead of as a debt, no legal repayment obligation with more flexibility than traditional debt — albeit Newtek plans to fulfill on its promises.

In summary: common stock now represents 100% of stock (post full-dilution), stock structure here is straightforward, with Series A, now sacrificed and Series B preferred serving mainly as a capital raising method with no equity dilutive ownership purposes.

Liquidity: Newtek trades 200,000 to 250,000 shares per day on average, at an avg price of $12 equates to $2.4million to $3million, daily volume. This is relatively low for a ~$300million market cap company.

10. Growth Runway

Loan Growth:

While Newtek is among the larger players in SBA lending, there is still significant runway ahead considering ~$31.5 billion, of loans were originated in 2024, for SBA 7a, of which is expected to continue growing at 3% CAGR. Whereas Newtek is targeting $1billion in originations in 2025, for SBA 7a, meaning market share is ~3.2%. This doesn’t consider other segments such as traditional commercial real estate (CRE) loans, is such a vast market that Newtek, doesn’t even make a dent, at its current size.

The limiting factor for Newtek’s loan growth has come from: primarily the growth of the amount of referrals being generated through the referral program, and the pace at which Newtek can continue growing their partners in this program. Secondly the access to available capital, though this is decreasing as a limiting factor as Newtek’s deposit growth outpaces loan origination growth alongside credit facilities in place to be able to sell loan portfolio’s to investors to cycle through capital, there are still some capital requirements which are needed to meet basil requirements.

Deposit Growth

Whilst deposits do make up 53% of capital, there is still a long runway ahead, considering: Deposits can get to 100% of capital funding whilst lending is growing at c.17% per year, to get to a similar position of LiveOak bank.

One may think, once deposits grow to 100% of capital funding, then deposits growth would be limited. However in Newtek’s case in particular, would still be able to use additional capital from deposits to hold a larger proportion of loans on balance sheet, rather being required to sell loans at lower rates (though still high ROCE) allowing Newtek, to net more by holding the loans — though it is worth pointing out once this point is met, the incremental profitability per additional deposit amount would decrease (as the investors capital is ‘less expensive’ if its being sold at a 6.5% yield compared to prior, borrowed capital from Deutsche Bank and CapitalOne at 9.5%).

Overall I do see room for Newtek to continue growing revenue near projected rates of ~17% CAGR, with Net Income growing significantly faster as low-cost deposits are becoming a larger share of capital funding.

11. Risk & Downside Protection

The following are the main risks & perceived risks I see for Newtek, of course, like every business Newtek has its own risks.

SBA Program Reliance, (~60% of loan revenues) rationale for SBA program highly likely to remain long term: Seen as public private partnership, with ability to indirect boost employment, program has received Bi-partisan support over many political parties, over 70+ years since being started, nor has there been any discussion to change.

Even in a scenario where there is a significant restriction/removal of SBA program, while there would be an obvious immediate impact. Newtek would still be primed to transition, as there is already a significant underserved market for SME lending which would only worsen if SBA program were removed, meaning Newtek would be able to take advantage of this and scale it’s Alternative Loan Program, which targets the same market but isn’t SBA reliant. Additionally Newtek would still be well positioned to transition to some other form of SME lending structure considering its decade of experience in the space, alongside it’s access to low cost of deposits.

Fraud/Numbers are Too Good to be True? Considering Newtek has paid $10.80 back in dividends (dilution-adjusted) which is nearly the entire market cap in the last decade, to me significantly reduces the likelihood of any potential fraud, for anyone who may be thinking the numbers are too good to be true. This is also seen by how their margins have followed peers during periods where Newtek had been affected by the same headwinds causing margins to drop, whereas a dishonest management would adjust numbers to make margins seem more consistent which is often not reflective of how volatile the reality of business, is — for me I am not seeing these traits occur.

Bank Run on Deposits: Do not see this to be likely considering, Newtek has been extremely cautious of which clients are taken on in terms of deposits, by eliminating significant depositors from entities who could be likely to shift capital to areas with higher yields (causing downfall of SVB). Consistently offering competitive rates to depositors savings accounts, preventing risk of capital flight, whilst also limiting % of deposits in a checking account, which has a much lower interest rate of c.1%, this is only about ~20% — as again Newtek doesn’t want to risk retention issues on its deposit base particularly in scenarios where yield starts to become more attractive elsewhere as rates had risen, nor is the business under any disillusions, that their consumers will purely stay for their brand.

As highlighted in the Financials chapter where I had talked about capital liquidity, the business, has enough liquid capital to more than likely fare, even if, in the unlikely situation, deposits were to drop significantly & sharply. Wouldn’t even see this likely to occur considering 75% of total deposit volumes is covered within FDIC insured limit.

Gain On Sale Is Unsustainable: This has been a re-occurring transaction which is a core part of the businesses net income, unlike that of regular banks. This has more than a decade track record of consistently occurring, additionally this will be occurring more frequently as the loan origination volumes keep increasing — cycling through this capital ~3x per year.

Even without this Newtek, would still do well, holding their high yielding loans, against their fairly low capital mix made up of 53%, deposits at ~3.83% and other capital primarily through Deutsche Bank & CapitalOne at ~9%.

What Would Actually be a Threat? Here I take quite a cynical view, as I want to see what would need to occur for the business &/or stock to suffer long term, to understand risk factors at a deeper level.

Newtek, is counter-cyclical mostly benefits from recessions (due to significantly lower competition as traditional banks halt business lending during such periods leading to higher growth rates). Here’s the most realistic scenario, I see, which could permanently cause a capital loss & requires the following to occur simultaneously:

- Sudden sharp raise in rates — though this would only harm Newtek if their loans reverted back to a primarily fixed offering (which I do not see occurring for at least the next few years, as the industry has only recently shifted to variable loans as primary offering, to avoid taking losses in this very same scenario.)

- An extreme recession, which would cause high default rates, this in itself not being an issue considering the strong LTVs of 70% to 75% being more than enough to offset the outstanding loan amounts. However defaults, cause for recourse on collatoralised debt to be actioned.

- As LTVs are strong, realistically asset prices, mainly commercial property, would be required to fall (45% to 50%+) over a short span of time to start taking any serious losses. To give an idea of the likelihood of this occurring, even in the 2008 financial crisis, where housing prices collapsed, on average commercial property prices fell (~35%) from the peak.

All three of these factors would need to occur simultaneously for any sustained impact to/jeopardise Newtek. Is this possible? Yes. Is this likely? No. This to me, situation highlights the durability of Newtek’s business model, particularly in comparison to other lending peers, who are more likely to falter before Newtek, by then which would constrict supply further.

12. Valuation

& Why book value doesn’t work?

I will caveat this segment by stating my own opinion on forecasting valuations;

The risk / reward should be so obvious that calculating the exact numbers shouldn’t be relevant, if it fails to be obvious, it’s a Pass.

That said I will briefly go over some Napkin Math’s to give an idea of the numbers:

Net Income & Multiple valuation:

2024 N.I — $49.25million

$49.25 x 1.20 x 1.20 = $70.92 million, 2026 YE Net Income (after 20% N.I growth YOY)

15x P/E

$70.92mil x 15 = $1,063.8million, 2027 Valuation

Current Market Cap = $311.0 million

Gain upon Recognition: +242%

Reasoning:

Net Income Growth: Clear line of sight, to 16% revenue growth over the next 2 years. Realistically 25%+ net income growth from revenue growth, increasing deposits as source of capital, continued reduction on loan losses on SBA as new loans price at Prime +3%. It is also worth pointing out that halfway through the year the business has grown N.I 25.88%+ however have used 20% N.I growth, to be conservative.

P/E Multiple: Multi Year growth ahead of net income 20%++, sustainable cost advantage, fairly wide NIMs of 3% acting as buffer, bank - so there is some bank risk, Somewhat diversified with high margin business services segment as well as High ROCE loan wholesaling segment with a long sustained 20%+ net income growth runway ahead. Easily command 20x P/E if not higher, however have decided to calculate at 15x P/E, for context LiveOak, a slower growing peer is trading at a forward P/E of 15x. Again, conservative.

Buybacks?
Worth mentioning, from my call with management, that they have considered buybacks especially at current valuations but ultimately are capital restrained and investing back into the business at 24% ROTCE makes sense.

I suspect if ROTCE were to potentially drop in the future (as the business scales) & if Newtek trades at similar valuations, wouldn’t be surprised to serious buybacks occurring.

Book Value & Why it doesn’t work here?

Book value is used as a banks measure of worth, by calculating based on the equity the business has grown and compounded over time, analysts then comparing banks on a price-to-book basis.

Going back to Newtek’s history, 8 of 10 last years have been as part of a BDC where 90%+ of earnings were paid as dividends as a BDC structural requirement, the book value has artificially been suppressed. Since 2015, shareholders received about $9.80/share in dividends (dilution-adjusted), versus current book of $8.65/share and a stock price of $12.03. I’m making this point to say, simply sticking a 1x book value, and comparing to a traditional bank which has been able to retain and compound capital on its book is short sighted — in this instance

Doing this is akin to judging a Car based on the paint job rather than the powerful engine hidden below. Newtek’s ability to generate income over time is just as important and explains how there has been a sustained ROTCE of 24%+.

Worst Case

Business currently trades at 4.8x p/e ratio, with a ~7.1% dividend yield. Realistic worst case scenario I see; the business continues trading at current multiples, whilst the business grows its is book value and pays out a growing ~7.1% yield. Without any capital appreciation, from purely dividend gains, this investment would near match the S&P500 historic returns. Wouldn’t see, even lower multiples to be sustainable, when considering management have weighed up share buybacks at current valuations, never mind at even lower valuations.

Though I even find this to be unreasonable long term, as the earnings power of Newtek continues to compound over time, resulting in a continually reducing P/E multiple.

Subscribe now

The Opportunity;

Valuation: 10x trailing earnings, 7x forward
Returns: 33.6% average ROE historically
Growth runway: <1% market share in trillion-dollar lending market
International expansion: Proven track record in US, Canada, UK

Competitive advantage: AI-powered underwriting uses granular data to identify creditworthy borrowers rejected by traditional lenders who rely solely on FICO scores. Loan sizes too small for manual underwriting make this economically unviable for such competitors.

Operationally Sound:
- Founder-led (~50% ownership), long-tenured executives
- Consistent profitability across economic cycles
- Clean balance sheet: current assets exceed all debt
- Counter-cyclical: credit risk improves in recessions as borrowers become more selective

Why I Passed: Incentive Misalignment

This business serves consumers with poor credit management at APRs averaging 110% (ranging 65-200%).

While it may be the best option available to these borrowers, the incentive structure concerns me. The business profits most when customers remain in debt long enough to pay substantial interest, but not so much that they default. There’s little incentive to help customers improve their financial position.

Yes, this critique applies to many lending businesses. But the interest rate differential here makes the incentive particularly strong—and unlike traditional lenders, there’s less concern about maintaining borrowers above certain credit thresholds.

Does this make management unethical? No. But as Charlie Munger said:

“You want to go for the win-wins. You want to look for things where everybody wins.”

I prefer businesses where shareholder value and management incentives actually align with customers.

The Alternative: SBA Lending at 4.6x Earnings

Valuation: 4.6x trailing earnings, 4.6x forward

The flat forward multiple reflects near-term quarterly headwinds and conservative analyst estimates (in my opinion, underestimating impact by switching capital funding source). But longer-term fundamentals remain strong: 15-17% revenue growth runway with net income growing faster. Short-term noise shouldn’t obscure the structural opportunity.

Returns: 17.7% ROE currently, improving toward 20%+ as higher-return segments scale faster

Growth runway: Despite being a larger player in SBA lending (still taking share), their alternative lending segment represents a fraction of the $2 trillion alternative lending market & expected to grow originations at a rapid pace. Margins expanding at an even faster rate, as a change in capital funding source reduces cost of capital significantly.

Competitive advantage: Significantly lower cost of capital than 90% of SBA lending peers. Operates an ecosystem (banking + payments + services) that reduces overall customer acquisition costs and improves long-term retention.

Also Operationally Sound:
- Founder-led (~5% CEO ownership). Track record of making short-term painful decisions for long-term benefit (proving alignment)
- Consistent profitability. Historically paid 6-8% annual dividends net of dilution, demonstrating financial resilience. Current yield: 7.9%
- Clean balance sheet: current assets exceed all debt
- Counter-cyclical: credit risk improves in recessions on corporate lending side, without requiring high interest rates

What’s Different:
Yes, both businesses lend to underserved markets and benefit from economic uncertainty & founder-led.

But the incentive structures diverge fundamentally:

Former: Average loan APR ~110%. Profits from consumers in financial distress maintaining debt loads.

Latter: Average loan interest ~12%. Profits when small businesses succeed and grow, requiring more capital and services over time. Higher rates than traditional banks, but in exchange for speed & flexibility by way of no restrictive covenants— not because borrowers have no alternatives.

One business model is incentivised by customers to staying difficult financial positions. The other benefits when customers spend on growth.

That’s the difference that matters to me.

Index:

1. Introduction & Deep History

2. Reason for mispricing

3. Payments

4. Marketplace

5. Fintech (loan, deposits)

6. Other Services

7. True Moat

8. Acquisition Rationale

9. Financials

10. Management

11. Shareholders and Liquidity

12. Growth Runway

13. Risks

14. Valuation

    Subscribe for more investing Deep Dives

Introduction & Deep History

This business is a true Super-App Ecosystem, one of the worlds best at that. To understand the business we will start with it’s humble beginning’s:

Originally formed in 1991, a venture between Arabian & Kazakhstani interests, months before Kazakhstan declared independence from the Soviet Union, as the country transitioned to a free market system, the banking sector was in its infancy.

In 1993 & 1997, banks had to re-register and meet new regulation, as part of this the name was changed to International Bank ‘Al Baraka Kazakhstan’ and then again to Kaspiyskiy Bank (name deriving reference from the Caspian Sea).

During this period, it operated as a small mediocre commercial bank, as one of many players in a fragmented market of undercapitalised and poorly governed banks. As the 1998 Russian crisis hit, this impacted Kazakhstan too and as a result several banks failed, Kaspiyskiy Bank survived though its share remained small, whilst large players started to consolidate & dominate, the likes of Kazkommertsbank, Halyk Bank, and BTA Bank.

2002, Vyacheslav Kim, acquired the bank as part of his vision to integrate the financing offering into his main business ‘Planeta Elektroniki’ (Electronics Retail Stores). However later stating:

‘At the time, successful entrepreneurs were buying banks, so we followed suit. But we did not understand how the banking business operated.’

This led to the critical point, in 2004, of Kim selling a controlling 51% stake to Baring, a private equity fund started by a U.S investor, known for successfully investing in tech enabled companies in post-soviet markets such as in Yandex.

Following the acquisition the business was rebranded to Kaspi Bank. Importantly Mikheil Lomtadze, rising star of Baring Vostok, had been appointed CEO at 31, in 2007.

During this time most banks had been focusing on corporate lending, expanding branches, traditional services & ignoring the digital shift occurring internationally and had worked well as Kazak economy was still booming from 2007 sky-high oil prices.

Short into his tenue, still in 2007, Mikheil, took like what seemed a completely irrational approach and shifted away from corporate banking to consumer financial services, in its entirety & quickly! Seeming like a crazy decision at the time as the business exited some of its largest relationships and purposely shrunk.

This worked massively in Kaspi’s favour as 2008 financial crisis hit, particularly by then mid-way through the year oil prices collapsed and so did much of the Kazak economy due to oil dependance alongside heavy borrowing in foreign currencies to fund expansion, which ultimately froze when the Tenge devalued significantly. BTA Bank, the largest bank collapsed with the help of fraud & significant non-performing loans, later rescued by government.

Alliance Bank, Temirbank, and others had then similar outcomes, as many banks had non-performing loans exceeding 30% of total loans, many depositors being burned in the process & the government having to spend billions to save & nationalise several banks to prevent complete collapse, though a prolonged process.

Kaspi being small & conservative allowed it to survive independently. Though loan quality deteriorated and profitability sunk—Mikheil’s call to exit risky corporate lending and focus on smaller consumer loans meant Kaspi wasn’t overleveraged in foreign currency, nor exposed to speculative CRE lending, which killed others.

Kazakhstan’s banking sector was broken, public trust was destroyed, whilst mobile adoption was accelerating, internet connectivity was improving, and younger Kazakhs were increasingly comfortable with using tech.

Traditional banking model—physical branches, relationship managers, paper-based processes had failed.

Lomtadze & Kim saw — digital-first approach could rebuild trust while offering superior convenience and lower cost structure.

Time Period 2010-2012, whilst the industry was attempting to recoup and recover, Kaspi invested into online infrastructure, developing its mobile app, digital payment systems & credit rating algorithms however still maintaining some physical branches, as cash was still predominantly used at the time.

2012 first version of app released, basic features - balance checks, simple transfers, bill payments. But a first of its kind in Kazakhstan, where in person branch transactions & unreliable ATMs were the norm at the time, unlike Kaspi’s app which was consistent & convenient.

SuperApp Buildout, 2012-2016, Conceptual shift: Rather than viewing Kaspi as a bank with digital features, it was now viewed as a technology platform that happened to have a banking license.

This distinction was fundamental. A bank with digital features optimises for traditional banking metrics—net interest margin, loan growth, deposit market share. Kaspi instead optimised for user engagement, transaction volume, and created ecosystem stickiness as per the below.
In 2013-2014, Three interconnected platforms were built:

1. Payments Platform — paying for utilities, mobile phone top-ups, internet bills, TV subscriptions, and government fines/taxes through the mobile app. Kaspi systematically added utility companies and service providers, creating a network where users could handle virtually all recurring payments digitally. Kazakhs who had never used online banking downloaded Kaspi because they needed it to pay their electric bill in a convenient & efficient manner. By 2016, Kaspi was processing the majority of utility payments in major cities.

2. Marketplace Platform — Created an integrated experience where consumers could browse electronics etc, compare prices, own instant financing, own payment system, and track delivery—all within a single app. Kim’s retail background aided, as Kaspi understood merchant & consumer needs.
   More consumers attracted more merchants, and cycle continued. Each transaction refined its credit algorithms, improving lending returns. Growing scale let Kaspi secure more suppliers, better supplier terms, lower prices, and faster adoption.

3. Fintech — provided instant loans, deposits, etc all digitally. Loan applications were approved in seconds based on credit scoring. Kaspi moved away from:

   Traditional underwriting — credit history & collateral.

   Instant approval approach — algorithmic models using alternative data—payment behavior, app usage patterns, marketplace activity.

This created stickiness which couldn’t be matched by other siloed banks, app becoming daily use infrastructure hence the massive uptick in transaction volumes.. This lead continued as Kaspi’s flywheel effect reinforced itself with the business growing form 3mil active users in 2016 to 8+mil in 2019 (nearly half of Kazak population of 20million).

By 2019, Kaspi processed over 80%+ all Kazakhstan utility bill payments. Alongside mobile service, internet, government fees etc. Marketplace platform grew aggressively, with “buy now pay later” (BNPL) being popular. While the Fintech platform held 8+million depositors, which is 40%+ market share making it the largest in the country, alongside a focus on SME lending & consumer loans where the credit scoring algorithm takes advantage of non-traditional data available exclusively to Kaspi.

2020, Targeted to cash out in part through a co-listing on the London Stock Exchange for additional liquidity, however this was originally blocked in 2019, due to concerns from Politically Exposed Person, the year 2020, listing being approved as the PEP individual had sold — deeper dive on this in ‘management’ section. Baring Vostok, Lomtadze and Kim all offloading some portion, while each party still retaining 20%+ ownership each.

Whilst Covid forced digital adoption, with huge benefit, this can be seen with the Marketplace segment growing 100% YOY with payments growing at a similar rate.

2021, launched Kaspi Travel, airline/train tickets & holiday packages — leveraging their ecosystem to become the No.1 in online travel in Kazakhstan in 18months!

2023, Kaspi entered auto market via acquiring 51% voting rights in Kolesa, which Baring Vostok, divested to Kaspi. Platform includes: browsing car listings, finance purchases, registering ownership transfers via government-services in the app, and are moving towards selling cars outright (both 1P and 3P).

Additionally acquired 90% of Magnum E-commerce, the online arm of Magnum Cash & Carry, investing ~$150 mil to build out grocery logistics and delivery. Magnum retained 10% and remains a strategic partner and primary supplier, providing: product sourcing, warehousing links, and retail infrastructure. Kaspi handles technology, payments, and fulfillment, while Magnum supplies inventory and physical store support— this operation became profitable within 1 year of acquisition.

During all this Kaspi, still remained customer focused, e.g Mikheil mentioning in an interview that the business was surveying 300,000+ customers monthly, taking on feedback.

2024, uplisted in the U.S via Nasdaq in January before shortly delisting in March from London, as there is far superior liquidity in the U.S. While a Short-seller report came out attacking Kaspi’s ties to Russia among other accusations (which will address throughout), Kaspi later refuted and confirmed only 2.8% of sales came internationally.

October, Kaspi acquired 65.4% stake in Hepsiburada, who is near break-even & Turkeys’ largest e-commerce marketplace for ~$1.12billion with 11.5million customers out of a potential 85million people market, crucial towards reaching Kaspi’s hitting 100million user target.

2025, Acquired Rabobank A.Ş, (a subsidiary with almost no operations) for it’s full Turkish Banking license, as part of the plan to replicate the Super App model: integrating Hepsi’s marketplace with new payment and fintech platforms built on Rabobank license. If successful, could transform Hepsiburada from a money-losing standalone e-commerce site into a profitable ecosystem.

While Turkey has its own struggles with hyper inflation (60% in recent years) & devaluation of Lira, alongside it, this has started to stabilise in recent years. As well as increased competition from AliBaba’s, Trendyol. Kaspi has made other acquisitions previously, though more in Acquisitions segment.

Additionally it is worth mentioning, throughout Kaspi’s journey, capital funding has always been fairly difficult to obtain, hence the business operating has operated with the mindset of, any new ventures must be: nearly instantly profitable (or at least within a year), and has contributed to an ROIC ranging generally from 40% to 80% historically.

Reason for Mispricing

Why is such a business only trading at 7x earnings to begin with?

Location Location Location:

Kaspi, located in Kazakhstan, where the equity markets aren’t as developed, being remote from the west certainly doesn’t help.

This business has had limited international exposure until recently as having been listed KASE exchange & LSE exchange in October 2020. The business has only recently listed on Nasdaq, January 2024 (via ADS) then delisting off London shortly thereafter in March 2024, this was due to the exposure and liquidity provided in the LSE being insufficient. Now that the business has been public in Nasdaq for more than a year I’m expecting exposure to start picking up.

Kazakhstan itself has fundamentally higher risks, which I will cover in the Risks segment.

Short Report:

Announced in September 2024 attacking Kaspi for its ties to Russian consumers among other things, while the peak of this seems to be largely over, its still worth mentioning as it did negatively impact performance/trust as the business fell in value over 20% overnight. Kaspi has since refuted & confirmed international sales made up 2.8% of total sales. Alongside a safety measure of all non-Kazakh residents requiring to open a card in-branch rather than online.

Whilst some of the points mentioned in the short seller report are certainly valid, they will again be covered throughout and can be found in the management & shareholders’ section.

Dividend Cut:

As most know, dividend cuts are never popular, personally I am in agreement with the decisions taken to cut the dividend as this was used to fund the acquisition of Hepsiburada alongside taking out some debt which being paid off. It had always been made clear to investors that the dividend is used to distribute cash that is in excess.

This can be seen, as on the date of announcing management are voting against a dividend on the 24th Feb, marked by a recent peak for the year.

iPhone Sales Hit: A month later March, 2025, Kazakhstan has set regulations in place so every phone needs to be officially registered and verified. This killed the gray market, shutting down people buying cheaper phones from unofficial sources, as a result Marketplace sales dropped 7% overnight, short term hindering growth estimates and negatively impacting short term sentiment, which is meaningless as a long term investor.

Revised Sales Target: Guidance has been lowered to an average of 15-25% guidance as phones sales hit, new capital lending requirements and marketplace with limited user growth have all seen a slowdown — I believe this to be due to a mix of short term issues as well as underlying growth slowdown, some increased competition (mentioned in MOAT) is one of these, however in summary this is more to do with temporary heavy advertising promotions in industry.

Whilst increased competition is never great, I do not think this is a serious threat for Kaspi, particularly as the business is very proactive in relentlessly strengthening its offering for the customer, my contact in Kazakhstan has confirmed that users can tend to temporarily use other offerings when given free incentives, e.g 50% off grocery purchases etc but generally no enduring impacts.

Payments

As outlined in the intro, the payments segment really started taking off when utilised for the payment of utilities, which was a new convenient method of payment. By introducing this into the ecosystem, usage internally such as purchases within marketplace etc became frictionless once integrated.

The market share growth didn’t come without any competition, Visa & MasterCard the existing giants, had already been dominating the Kazakhstan market, their MOAT, globally is known for being a tough nut to crack.

The image below really highlights the strength of Kaspi, to quite frankly be able to wipe the floor with the incumbents, in a singular year to be able to flip the market share completely in terms of total monthly Processed Volumes in Kazakhstan:

Source: Page 83, 2020 London IPO Prospectus

What does Payment transfers look like on a daily basis?

1. Kaspi QR In-Store Payment (~86% of transactions) Tap “Scan QR” → Point at merchant’s code → Authenticate → Done in 5-8 seconds. Completely Instant settlement.

2. P2P Instant Transfer Enter phone number → Amount → Authenticate → Money appears instantly in recipient’s account. Free, no limits.Common use case: splitting bills, paying friends/family.

3. Bill Payments & Utilities (~84% share of utility bills)
Select utility → System pulls your bill → Pay → Done in 10 seconds. Electric/government databases update instantly. Free for consumer.

Why These Dominate: All free, instant, phone-only, no cards needed. Drives 4-6 daily app opens and makes Kaspi essential infrastructure & provides significant data which then aids in its auto-underwriting process.

It’s also worth mentioning there has been a payment partnership AliPay & Kaspi recently which is immensely beneficial as there is a decent amount of tourism from China into Kazakhstan, whilst the exact split on payments hasn’t been announced. This at a minimum opens up the ecosystem benefitting merchants and likely to some extent being an additional revenue driver on payments.

Marketplace

The above provides a pretty good round up of where the marketplace business is heading. Even with a change in regulations, now more strictly monitoring & requiring iPhone authentication causing an estimated 7% sales slowdown impact.

The breakdown shows the frequency of purchases increasing 35% whereas active consumers have only grown 10% highlighting that this is the real bottleneck for the business.
Kazakhstan has around 20million people meaning, based on best-available estimates around 14million (or 70%) of people in Kazakhstan would be adults who would actually be able to use the platform. Of which 8.4mil users / 14mil adults in Kazakhstan = 60% market share (if not higher).

Realistically this means there will be limited growth for user growth to grow meaningfully beyond population growth of 1.2% & growth from late-adopters in Kazakhstan, though there is currently still room for growth to come from more of the current users using more features over time, such as e-Grocery, which is still fairly new, as e-Commerce GMV are currently split: (for 1H 2025)

1. e-Grocery — 6%

2. e-Cars — 28%

3. General Goods — 66%

Growth is therefore driven by frequency (which is tied to more overall usage of entire platform offering) which has been still been increasing at a healthy rate when looking at frequency of purchases increasing 35% YOY.

Though it is worth pointing out, average ticket size, a previously important metric, has been purposely de-emphasized by management as average ticket has generally decreased, as per recent vague commentary and we as investors are no longer told the amount (nor is sufficient data provided to calculate).

However to me, this is nothing to necessarily be concerned about, as the marketplace mix is generally shifting, e-Grocery, the newer venture is growing significantly faster, even though there is a lower average ticket price and margins, this is not a concern as the unit economics per item aren’t as high. E-Grocery businesses work off high-volume, low margins.

Additionally Kaspi no longer discloses it’s market share —why?

This usually happens for 2 reasons:

1. Market share growth has been unimpressive & shows weakness so its hidden

2. Much less common, the business does so well, that talking about it can be counteractive because it would blatantly reveal a monopoly and raise unwanted attention.

Kaspi’s Market share of marketplace is that of a monopoly, estimated ~65%, as a result talking about it in the open has more to lose than gain, as the saying goes the tallest tree is first to get struck down by the wind.

Therefore for me, it makes complete sense that the business has stopped reporting on this metric in the last few years.

Postomats:

Kaspi has 8,000+ postomats, automated parcel lockers, acting much the same way as the western equivalents: Amazon Lockers & InPost Lockers.

To give an idea of the scale of the postal network via comparison.
Kaspi has 8,000+, whilst the largest competitors only have hundreds, not thousands!
This network of postomats allows the business to increase convenience of pickup & reduce last mile delivery costs, as this avoids multiple failed delivery attempts/consumers having to be restricted in their homes to wait for parcel deliveries.

Whilst the network isn’t exclusively used by Kaspi, operations are limited to approved postal partners & marketplace sellers. Another important factor for Kaspi’s marketplace business as competing non-Kaspi E-commerce firms aren’t granted access, whilst competitors might still partner up with other businesses with Postomats offerings the density is a vastly inferior offering because of this.

Fintech

Loans:

Loan book split:

BNPL — 40%, is the largest segment which has scaled along accordingly with growth in the Merchant offering, works by charging the merchant a fee for offering e.g instalments for up to 180 day period. This making sense for the merchant as, is willing to trade a few points on the sale, to aid with conversion.

General Purpose— 37%, General SME loans to their consumers, these are auto-underwritten and have instant approval compared to other offerings in the market.

Merchant & Micro Business Finance — 18%, as you will notice a large portion of these loans grow through integration with the rest of Kaspi’s platform, Merchant & SME loans are no different and help Merchants, largely who operate on marketplace.

Car Finance —5%, Again likely a large portion of leads come from Kolesa, Car classifieds marketplace.

Positioning on Loans: Kaspi, charges a more premium rate, which it is able to do mostly as a result of the convenient experience in platform, when considering a significant portion of the purchases occur within the platform alongside the fact the loan process itself is: seamless, no in-person visits required, loan is approved instantly through the use of auto-underwriting (compared to traditional manual underwriting).

Kaspi is able to more profitably underwrite these loans due to significant data the ecosystems provides such as account purchase history, deposit history, visibility on cost to income ratios, cash balances, length of account being open, this allows Kaspi to access non-traditional data exclusive to itself, which is used to Auto-underwrite, cutting manual costs & reducing any delay. As a result default rates are significantly lower than competitors.


Double the data, half the default rate:

This can be observed when comparing Kaspi consumer loan delinquency rates to that of competitors, Kaspi having averaged 2.1% whereas Halyk bank, a titan bank in Kazakhstan had a 4.5% delinquency rate, as per their 2024, Annual Report, pg40. Whereas ForteBank can be calculated at 3.61% for 30 day delinquency as per the most recent available 2023 English Annual Report, pg124.

This gives us a real idea for the average, post 30 day delinquency rates in Kazakhstan and shows there is tangible benefit in Kaspi’s extensive and most importantly clean dataset which it benefits from, when it comes to it’s auto-underwriting approach.

Fixed/Variable loan split:
Essentially loans are offered are pretty much only on a fixed rate basis, so there is some interest rate exposure however for me I am not too concerned when considering BNPL (40% total loans) & likely a decent portion of General loans (accounting for 37% of total), of these loans, would have a less than 1 year in loan duration therefore changes in interest rates, on this portion having minimal impact compared to longer dated loans.

Even when considering the longer dated loans, Kazakhstan base rates had already been at 16.5%, increased rates are due to managing inflation, which seems to be working so wouldn’t necessarily expect to see any significant increase, even if it did occur Kaspi has strong NIMs on average of 7%, giving a lot of cushion if base rates were to rise.

Loan Volumes:
Volumes has in effect been flat for the year with:
1H 2024, 6.4million active loans VS 1H 2025, 6.3million active loans
This is driven by:
- increased regulations on BNPL
- Smartphone sales down following new rules
- Comparative headwinds as rates this time last year had fallen VS currently rates have increased.
In my opinion, this segment is starting to mature in terms of volumes, for Kazakhstan, given that Kaspi holds significant share, if anything would expect growth to, track inline with marketplace for me this is not an issue as this is not where I would be expecting major growth for the business.

Deposits Breakdown:

Deposits are split between 85% term deposits, which is higher than in most countries where you often see banks with more sizeable portion of current accounts, in terms of deposit split. However in Kazakhstan, where currency devaluation is higher & interest rates are significant, consumers would be significantly penalised for keeping their capital in a low yielding current account, therefore not too concerned with this split, as it is to be expected in Kazakhstan.

Focusing in on the deposits, Kaspi has recently launched its new term deposit offering, after the Kazakh base rates (very recently) increased to 18%, this product has had significant traction, so whilst this has obviously cannibalised the previous offering of rates of about 15% to 16%, reasoning behind has been:
- Needed to remain competitive following rate increase.
- Of the customers who had signed up for this term deposit, 69% were existing inflows & importantly 31% are new inflows & users to Kaspi.
This ended up working successfully as the offering went from 44K depositors in Q1 ‘25 to 159K users in Q2 ‘25, a 263% increase QoQ, this has largely driven overall deposit growth this year of 19% YoY.

Overall considering Kaspi, is still able to maintain a NIM of 7%, shows the strength of the business model, which in part comes from being able to charge a premium on loans as well as often having slightly lower yields, which they are able to do given the convenience & stickiness of the ecosystem.

Term deposits provide safety, in eliminating liquidity risk, which is also why, for me it is a strong sign to see that the vast majority of depositors are consumers, who will stay within the ecosystem, unlike Merchant/Business deposits, only accounting for ~2.4% of deposits in ‘24 which, more often will, switch in pursuit of the highest yield.

Real Bank Run Test Scenario
In 2022, what seems to be the peak of Russia-Ukraine conflict, Kazakhstan had deposit withdrawal concerns: Sector outflows of 8-12% whilst Kaspi only saw a 4% dip in Q1 2022, mitigated by app alerts and incentives; full recovery by Q2, with 15% year-end growth, this proves the ‘stickiness’ is real alongside the business now being significantly more well diversified with two thirds or more of net income, coming outside of the banking segment. This places Kaspi in a position of strength if any
Bankrun scenario were to occur, even more so now, as Kaspi begins to diversify into a new economy (Türkiye, more below under: Acquisitions)

Other Services

A quick overview on services provided

Merchants:

1. Tools to accept payments via QR/cards/app

2. Calculating, filing, and payment of taxes

3. Kaspi Shopping Register: Payments in a POS system (this includes product Catalog, tax invoices, sales records).

4. Kaspi Restaurants: QR-based payments & tips for restaurant tables

5. Membership, to receive: Dashboard analytics, Inventory management, Product catalogue, Priority Support, Cashback offers.

6. Third-Party Integrations: Acceptance from other banks like recent Alipay collaboration (which is particularly beneficial for the use-case of Chinese tourists visiting Kazakhstan.)

7. Verified Devices (e.g., smartphones) for authenticity.

Consumers:

1. Household Bill Payments: Payment of utilities and recurring bills (60%+ country market share)

2. e-Grocery: Ordering and delivery (In 5 cities via dark stores) via Magnum, e-Grocery 90% of e-business.

3. e-Cars & Properties: Online vehicle sales (1P and 3P models) via Kolesa & Properties via Krisha.kz — Kaspi acquired 50.76% of Kolesa, in 2022 from its major shareholder Baring Vostok.

4. Kaspi Travel: Booking rail/air tickets, tours, and package holidays (domestic tours added 2025) — became the no.1 player in this market in 18months!

5. Logistics with direct-to-door/pick-up points including ~8,000 Postomats.

6. Kaspi Classifieds: Listings for used/new goods & services (think Facebook marketplace & Craigslist) , as well as Job listings (one of largest Job platforms in Kazakhstan typically mid/entry roles).

GovTech:

1. Storage and access of IDs, like passports, .

2. Driving License Renewal, Car Title Transfer, Paying Fines etc.

3. Legally registering businesses.

4. Tax Calculation and Payments: Frictionless Tools for tax computation and payment.

5. Simplified Submission of tax reports (unlike the incumbent government portal).

6. AI Virtual Assistant (Ruslan): Support with general queries (e.g relating to marketplace purchases.) — estimated to have saved ~2,000 customer service related jobs.

The Government services do not generate any income however acts as a significant lead magnet, with real use case where it greatly simplifies the tax-paying process. This has real utility & stickiness to Kaspi, and a difficult to replicate feature given this is the result of multi-year government integration.

In general as shown here in the points, these services are vast have real-world applicable use cases, which is part of what makes the ecosystem so valuable.

True Moat

To me, this is one of the rare businesses which has a true MOAT, which has led the business to have ~70% market share in Kazakhstan in the likes the of the marketplace & payments segment. On average over the last decade Kaspi has had a ROIC of 91.6% over the last 9 years! Though over the 6 years this had dropped to 64% on average, as it became harder to achieve as high returns with increased scale, and has been part of the reason why management had returned significant amount of net income back to investors via dividends - Regardless, ROIC is excellent, far beyond an average business at 10%.

So what is Kaspi’s MOAT?
The SuperApp ecosystem is the key aspect, which then has multiple advantages within, especially as they’ve scaled. However I need to acknowledge the business has only managed to get to this dominant position because of managements, customer obsession.
Mikheil believes in essence, if you do not delight your customer with your existing products then there is very little incentive to try any new product within the Kaspi Ecosystem, this is what has allowed them to scale so significantly with very little advertising needed, making Kaspi exceptionally capital efficient.

Moat Components:

1. Payments, benefit from Network effects, as more people pay with Kaspi, it becomes easier, for example: within your personal contacts, those who are Kaspi users are saved and can transfer usually in under 10 seconds, not minutes. As well as charging no/low fees on transfers, given Kaspi monetises off the whole ecosystem.

2. Government integration (which takes multiple years of collaboration to setup), acts as a significant lead magnet, as it is the only free service which significantly simplifies & speeds up: Tax paying process, Paying fines etc.

3. Significant Data advantage, particularly at the scale of almost the entire country allows the business to, receive insights into spending which can then be used to research; what segment to next enter into — such as Train/Flights booking platform.

4. Again Data, allows Kaspi to implement Auto-Underwriting, which has lower 30-day delinquency rates (2.1% whereas competitors around 4%) — saves on manual underwriting costs, instant approvals (Fast & Convenient). This is only possible by having multiple non-traditional data points such as: Account balance, Income & expense history, account inception date, buying habits, any other outstanding debts, track record of paying off loans etc.

5. 8,000+ Postomats, has effectively about 10x more Postomats than the 2nd largest player in Kazakhstan, which is exclusively used by itself and some non-competing partners, more comprehensive delivery/drop off options for Marketplace business.

6. Marketplace - Network effects at play, estimated 65% - 70% market share, more users attract more merchants, which improves offering & prices, better customer experience, more customers, virtuous cycle

7. Limited advertising costs needed to promote & expand new features, riding on the back of goodwill Kaspi has built up with its other features (79% brand awareness in payments), unlike competitors with supposed ecosystems, who need to outlay significant capital to attempt to attract customers away from Kaspi.

8. One of the least tangible but for me, most important, Kaspi’s Culture, focusing solely on Net Promoter Score targets, rather than sales targets.

9. High switching costs due to super-app integration: Convenient all-in-one app ecosystem locks in users, making it inconvenient to switch providers—evidenced by an increasing, 68% DAU/MAU ratio (2nd globally only to WeChat, a 1 Billion-user platform) & 73 monthly transactions per active consumer on Kaspi.

10. Economies of Scale: fairly common for larger businesses such as Kaspi, more pricing power, given the scale of operations.

The above, Shows Kaspi’s DAU/MAU, as well as the progression over time, as shown, this is currently amongst the best globally in industry. This has been increasing over time, whilst the rate of DAU/MAU increase has been slowing, it has still been increasing, not shown in the image is that Kaspi’s DAU/MAU had been:
Q1 2025 - 68%, Source: Pg 32 & Q2 2025 - 69%.

This to me highlights, that the business continues to get stickier over time & whilst it did seem that there had been a flatline in 2021 & 2022, I speculate this is more likely due to significant increase in new user growth during Covid, meaning a larger proportion of new users, who initially utilised the platform less frequently.

Payments:

Yes this image has already been shown under payments, however this most clearly demonstrates Kaspi’s MOAT.
As shown, it is not an exaggeration to say Kaspi, wiped the floor with Visa & MasterCard, known for their fortress like MOAT. What does this tell you about Kaspi’s positioning?

Kaspi’s success is the extent of being known as the unofficial, national bank of Kazakhstan. Success often attracts attention and increased competition particularly as those players start to clone some of the successful features of Kaspi.

This can be seen in the development of Halyk bank, Jusan Bank, Freedom Finance who have started to implement key features as broken down for the 3 companies as an example:

1. Halyk Bank - QR code payments, digital wallets (google/apple pay), cardless ATM withdrawals.

2. Jusan Bank - mobile-first banking app, instant transfers & payments, digital lending

3. Freedom Finance - through their Freedom Superapp also offer Instant transfers & payments, digital card/deposit opening, loan applications in-app, currency exchange, utility bill payments, combining banking & brokering.

Though still not seen at the level of Kaspi’s vastness of services offered, nor consistency of service quality (which Kaspi is said to have, a micro example highlighting this is; the support team having an average 2 minute pickup time on calls, as per discussions with users) there has certainly been a step up in offering from the rest of the market alongside more aggressive promotions, e.g 50% of frequent selected purchases upon setting up a Freedom digital card, in an attempt to integrate the user into their ecosystem to then; convert to brokering services.

Whilst growth has slowed slightly for Kaspi, largely attributable to one-off changes such as phone sale regulations, capital requirement changes on lending etc, I do believe that increased competition has slightly contributed. However Kaspi’s MOAT runs deep as touched on, also important to not forget Kaspi is already & remains the default option engrained within culture, much like Google — meaning for this to be overturned in any real way, the offering would have to be significantly better than Kaspi’s already great service, which is continuing to improve (rare to see among monopolies) to which all existing services in Kazakhstan are statistically inferior, though respectable compared to other offerings in different countries.

Studying More Developed Chinese Market:

If we look at the more developed, Chinese market for lessons in this space, originally AliBaba through AliPay had been the incumbent with 80% market share, in payments, then had been successfully penetrated by Tencents’ WeChat (initially having 10% share) was able to utilise its sole monopoly in messaging to cleverly integrate the Chinese custom, which is gifting through red envelopes, in an online format. This became viral as the effect of sharing red envelopes online in a frictionless way with family & friends grew payments from almost 0 to 300million+ users within 24months for Tencents’ WeChat, particularly as users kept sending envelopes to others by design. Currently Tencent’s, WeChat & Alibaba’s AliPay have 42% & 53% of market share respectively, whilst having defended against multiple strong businesses from other verticals who have also attempted to enter, none other have penetrated successfully into payments. It is worth mentioning, though not in payments, Byte Dance (TikTok) was able to break into the Live Video Ecommerce market through their existing monopoly on short-form videos on TikTok to pair & promote, in a way the other E-commerce players are unable.

When applying this to Kaspi, we are aware that:
Unique market positioning such as dominance over social messaging can be a strong ways to disrupt other existing markets, when leveraging existing audience & service to apply a unique experience.
When looking for similar silos in Kazakhstan there are no immediate ones from competitors which could be used against Kaspi. I do not count Instagram, TikTok etc, as I doubt this market is big enough to warrant their interest & even if they were to take this approach, Kazakhstan certainly is not the first priority to expand into.

Realistic Potential Silos are:

1. Halyk bank - dominance over pensions in Kazakhstan however this is rarely visited by users, which is important as frequency of usage is needed for consumers for a switch to be effective in my opinion. As well as Halyk itself being the incumbent who is slow to innovate by comparison to Kaspi.

2. Freedom Finance - Offers international stock trading, it is still fairly infrequently used & in my opinion the management focus is not customer centric enough to be able to compete with Kaspi.

3. Beeline, Kcell, Altel (major mobile operators) however this in itself is fairly fragmented - however could see a scenario in which new payments platform is pushed via data contracts, potentially offering more favourable terms however this wouldn’t likely have the same virality effect as experienced in China, nor would it be as successful given fragmented market.

In short, no obvious high frequency usage, high market share platforms, which are visible risks to disrupting Kaspi, meaning in general, well positioned to remain in dominant positioning, particularly as Kaspi is continually improving, as evidenced by metrics such as DAU/MAU.

Acquisition Rationale

Portmone Group (Ukraine):
Acquisition finally closing in October 2021 (for an undisclosed sum), the business had strong initial growth of 40% in 2021, after 2022 Russia-Ukraine conflict this segment of the business had been deprioritised, effectively on hold for obvious reasons — alongside intent to acquire a bank in Ukraine which had been approved by regulators but not pursued, later Portmone started flatlining (relative to prior growth).

Portmone, founded in 2002, holds Visa/Mastercard licenses, serves 2,500+ merchants, and processes UAH 15B+ annually in payments/transfers.
Reasoning behind acquisition at the time: Immediate market entry into Ukraine (population 40M+), leveraging Portmone’s 1M+ users for start of Kaspi’s super-app model buildout. Whilst the bank purchase had been approved, it was intended to buy the remaining component of the SuperApp (you will see later on this same playbook is being done in Türkiye currently).

Magnum E-commerce:
February 2023, acquired 90%, from Magnum Cash & Carry LLP. This is for the
E-commerce segment, the model worked by integrating e-grocery into the ecosystem which is an additional service Kaspi would be able to monetise, whilst for Magnum, this was an opportunity to offload an unprofitable segment of their business, which Kaspi, turned to profitable within a year, aided by the significant amount of users within the ecosystem. Magnum effectively keeping the remaining 10% E-commerce ownership as shmuck insurance, to benefit on the upside.

E-grocery segment is able to get attractive rates, similar to that of other Magnum stores, which Kaspi, would have not had the pricing power to do if started independently. E-grocery currently operates out of 5 Dark stores, within major cities within Kazakhstan & another 3 planned to open, with the segment having grown 50% YoY as of most recent. Note, from a numbers perspective this segment will cause for ‘drag’ on their metrics such as net margin percentages having been 40%+, whereas grocery being high volume, low margin, will drag this metric down, though it is still profitable, so a net benefit to Kaspi, but something to be aware of.

Kolesa Group (Kazakhstan):
Acquisition closed October 2023 for 39.7% acquired from Baring Vostok (private equity & major Kaspi shareholder) for ~$88.5M valued at ~10x earnings with net income growth of roughly 300% YOY (based on 1H 2023 earnings of 6.7Bn KZT), a bargain in my opinion. Some sources cite 50.76% (however to me it seems Mikheil had transferred the voting control on his ~11% stake which brought Kaspi’s voting control up to 51%).

Business was founded 2014, operates Kolesa.kz (autos) and Krisha.kz (real estate) classifieds, with 10M+ monthly users and 1M+ listings. By integrating the offering into the Kaspi ecosystem Kolesa, benefits from significant user exposure (pretty much the entire country) & Kaspi’s brand, which is likely why even though Baring Vostok sold, as they are significant shareholder of Kaspi, they would still continue to benefit from the integration.

Kaspi obviously benefitted too, this was a market leader in classifieds, another highly profitable service which could be integrated into the ecosystem, which was already proven to work independently, to me this is one of the rare acquisitions where the word ‘synergies’ could actually be used, for example Kaspi’s car loans up 50% post-deal!

Humo (attempted):
In 2024, Kaspi mentioned of wanting to acquire Humo, which is part of Uzbekistan’s domestic card payment duopoly, with an estimated share of 60%.
However this was ultimately not pursued, though not publicly confirmed I believe this was due to unease from the local government around a major foreign business potentially having a monopoly like ownership on the payments market and more particular if Kaspi were to expand its offering, particularly as one of the major reasons behind Humo, being setup in 2018 was to prevent reliance on a singular payment system & not wanting a monopoly market in the space.

On this basis the deal was likely prevented/not worth pursuing when considering the likely demands to satisfy the unease, though this is my personal speculation so shouldn’t be taken at face value, when assessing the purpose behind why Humo had been setup, beyond the vague answers given within IR.

Hepsiburada:
Kaspi acquired 65.4% of the E-commerce giant & closed by Jan 2025, as part of Kaspi’s expansion into Türkiye, consisting of 86million people.

Hepsiburada’s active customer base is ~11.6 mil, been flat/-2% over the last few quarters due to the mix of the following: Stricter metric definitions e.g Kaspi definition of GMV excluding returns/fees; generally tough market in terms of inflation which has hit Hepsi harder, than that of competitors mainly Trendyol (largest player in the market) who is frequently aggressively discounting unlike Hepsi, who I’m assuming based on management historic philosophy (typically targeting profitability within 1 year of a new venture) have cut significantly on largely unprofitable heavy discounting.

Whilst growth has certainly slowed profitability has improved:

However what is not shown here is, for the most recent few quarters, Hepsi has become unprofitable again (though only slightly, -1.5% operating margins), which I am not too concerned about given this is still the year the business is integrating with Kaspi & making heavier investments into their platform.

Competition in Turkish e-commerce is split as following estimates (based on best available reported figures):

As shown Trendyol, is the biggest player in the market with Hepsiburada, a fairly close second. Trendyol backed by Alibaba, operates solely through independent merchants though has grown quicker historically, I believe Kaspi’s timing to enter Trendyols’ main market has been great, as the business is currently expending significant capital to enter into CEE, Germany & Gulf countries, all simultaneously in 2025, which will inevitably place strain on the business.

Amazon & n11 are also notable players in this space however given their current share and rates of growth, am personally not too concerned.
My main thesis (& seemingly that of management team) is that focusing on the customer experience and integrating the ecosystem network, Hepsi should be able to achieve closer levels of stickiness as Kaspi in Kazakhstan.

The business has already started taking steps for this very purpose when assessing:
Hepsiburada already has the payments & marketplace element however by recently effectively acquiring Rabobank A.S for its banking license (though undisclosed, an immaterial small sum), Kaspi would be able to offer financing which will aid in creating the 3 core parts of the ecosystem.

Overall given the potential of this business for Kaspi, paying $1.1 billion, valued at ~1x revenue multiple (pro-rota for 65.4% of the business) to me seems worthwhile considering the potential Kaspi has to add value, even in terms of replicating successful elements.

Rabobank A.S
As mentioned above the acquisition of 100% of Rabobank A.S, is effectively a branch with limited operations being acquired for the purpose of its Turkish banking license, though the purchase amount was not disclosed I speculate this was a fairly small considering there had been no additional capital raise requirements & the amount of assets being ~$118mil USD (assets wouldn’t account for debt already on the balance sheet) therefore wouldn’t anticipate the purchase price to be too significant relative to Kaspi’s scale.

Purchase currently pending regulatory approvals from Türkiye’s BDDK and Kazakhstan’s ARDFM, targeted December 2025. Rationale: With the banking license; would be able to start implementing Kaspi model, by building out the Turkish ecosystem through combining services with Hepsiburada to meet the required 3 core services — Banking, Payments, Marketplace.

Acquisition Thoughts:
Overall, I believe Kaspi management have generated shareholder value, through savvy acquisitions which ACTUALLY play on their strengths (unlike lots of acquisitions which promise cost synergies) — I believe Mikheil’s’ P.E background & Baring Vostok, a successful PE firm being a major shareholder / having board members have contributed to this. Additionally from a risk perspective, the business has not overleveraged itself when acquiring and paying special attention to reducing debt, after large acquisitions such as that of Hepsiburada.

Generally acquisitions make sense for both parties as, Kaspi gets more favourable rates upon acquisition, if the seller is able to keep a portion of the original business , which it often does, knowing they will benefit from the additional upside in the lift from ecosystem integration. Kaspi itself obviously leverages the ecosystem model to boost original company earnings making the investments worthwhile whilst adding to the overall utility of the platform.

Important to note, a good amount of their successful acquisitions have come from the major shareholders directly, Kolesa was owned by Baring Vostok major shareholder, whilst Magnum has Vyacheslav Kim as board member & owner (though the amount ownership is not disclosed).

Financials

Revenue is well diversified within the business segments, with approximately a third of profitability, split between each segment as shown in the Q1 2025 breakdown. Though this is obviously concentrated within Kazakhstan, the business has limited from any exposure from Banking downturns, especially important as it protects Kaspi from any bank-run scenarios.

Income Statement:

Assessing the 1st Half comparisons:
- 55% Increase in revenue (17% increase when excluding for Hepsiburada acquisition, in this instance would take year-on-year comparisons such as COGS etc, with a grain of salt particularly as Hepsiburada is, though barely, still unprofitable).
- 37% increase in interest expenses YOY (for the 6 months), which is good considering interest income growth of 43.7% outpaced expenses growth.
- 353% increase in COGS, YOY (for the 6 months), however when looking at the most recent 3 month comparison, it is evident the significant increase upwards has been inherited from Hepsiburada post acquisition.
- 9.92% increase of profit YOY, again this isn’t ideal however considering Hepsiburada is still a small unprofitable drag, it is to be expected.

Overall, still solid profitability growth, in this instance I am not able to compare, like for like figures which masks, what is occurring to the business in terms of margins YOY.

Asset + Liability Breakdown:
Most of Kaspi’s liabilities are held as deposits which are used to fund lending, there is some USD debt exposure, often due to more attractive rates achievable then in the Kazakh market, as an investor I am not a fan as this would create significant currency risk for the business.

However in this particular instance the amount of debt outstanding in USD is $650million, through a Eurobond issued to investors, compared to annual net income of $2billion+, this is not a concern as payable within 4month period, which is what has been occurring hence the reason behind the dividend cut as Kaspi is serious on paying off this debt — on this basis do not see the debt as a real long term risk but will monitor payoff.

Source:
520.22455 KZT/USD on 30 Jun 2025
For original figures in KZT Open: Q2 2025 Financial statements, pg 3 & 5

Liability Breakdown:
As per the attached, it is not an exaggeration to say the vast majority of liabilities are arising from deposits & a smaller sum from the issued Eurobond as per the 2025 info available, the business does carry smaller amounts of other liabilities outside of these two, however this cumulatively only amounts to $1.88Billion relative to the $12.87Billion in deposits alone.

Asset Breakdown:

Cash&Cash Equivalents + Mandatory cash holdings + Bank amounts owed = $2.3bil
Effectively cash on hand available to use in business activity, whilst regulatory amounts held aren’t necessarily liquid this is a small amount held at only $86million.

Investment Securities & Derivatives = $2.3bil, this is mostly comprised of business ownership likes of 65.4% ownership of Hepsiburada, worth ~$969mil at current trading valuations. Ownership stake in Kolesa & Magnum E-grocery. Likely to also include the value of some interest rate / currency hedges, though no clear breakdown is provided.

Loans to customers = $12.28bil, similar to the deposits this is where the majority of assets are held, as you can workout, outside of the banking operation, the remainder of Kaspi’s business model is asset light.

Inventory = $214million, a fairly small amount.

I have intentionally not included the rest as I cant count on the remaining, in terms of intangibles, goodwill (which historically has been kept near zero, increased this year likely from Hepsi, acquisition. Other assets etc, frankly these are immaterial.

For me the main point is if: Kaspi can reliably and comfortably pay it’s debt.
What we know:
Total Debt = $15.14billion
Total ‘Actual’ Assets = $14.8billion (Total current assets = $2.3billion + Total loans receivable & Inventory = $12.5billion)
TTM Earnings = $2.1 billion

As shown in the numbers, sellable assets are sufficient to cover total liabilities in the long term, when considering Kaspi will continue to generate $2billion (and growing) per year in net income, see no issues fulfilling on their debt obligations long term.

Management

Management Background:

Mikheil Lomtadze – Co‑founder and CEO
He has led the company since its transformation from a small bank into Kazakhstan’s dominant fintech super‑app
Lomtadze earned a BBA from the European School of Management in Tbilisi and an MBA from Harvard Business School, of which during his time at Harvard mentioning to peers, his intention had always been; to build a business he would be proud to run & that people would thank him for (much like what occurs in Kazakhstan today with Kaspi).
Early in his career he co‑founded a Georgia-based audit firm (later acquired by Ernst & Young), then working at Baring Vostok where he was seen as a ‘rising star’ partner at the, private-equity firm. Lomtadze was brought into Kaspi, initially through Baring Vostok’s investment in Kaspi.
Kaspi is now seen as a National Pride, for elevating the countries’ technology.
He serves on Harvard’s advisory board for the Middle East & North Africa.

Pavel Mironov – Deputy Chairman of the Board & COO
Founding member of Kaspi’s management team (joining in 2008), he studied, Moscow Institute of Electronics and Mathematics, HSE.
His previous role was a computer scientist, previously worked on IT projects across the CIS at the European software firm Tieto.
At Kaspi he is responsible for daily operations & technology delivery across the platform including; development of Kaspi’s apps and back‑end systems. He mentions using “state-of-the-art technology” and data analytics to improve customer experience.
Kaspi has invested into their Directors to attend Harvard Business School executive program, of which Mironov and others on the list have done.

Yuri Didenko – Deputy Chairman of the Board (Capital Markets)
One of Kaspi’s first executives (joining in 2007), unsurprisingly has a finance background studying at Kyiv National Economic University, finance degree & CFA charterholder.
Previous role: Investment director at Baring Vostok. Current role focuses on treasury and IR. Played a key role in Kaspi’s IPO and bond issuances.

Tengiz Mosidze – Deputy Chairman of the Board & CFO
Joined the founding team in 2008 also has a finance background (MSc from the European School of Management, Georgia; also HBS GMP 2013). Before Kaspi he worked in Ernst & Young’s Central Asia finance group and on microfinance projects at the World Bank.
As CFO, oversees budgeting, capital structure, and reporting. (According to industry sources, he also oversees the Kaspi Marketplace platform.) worth mentioning his financial leadership has earned Kaspi praise among local business media and ratings reports.

Sergey Timokhin – Executive Director (Payments)
Has led the company’s payments service, Timokhin’s led, building: cashless payment apps (Kaspi Pay, QR Payments, etc.) that today are dominant in the country and revolutionary at the time, he has also driven product innovations such as palm‑print biometrics and in‑app loyalty features. Often representing Kaspi in regulatory or PR (for example, responding to regional payment inquiries) & in a 2018 interview he emphasized Kaspi’s customer‑centric strategy - being part of his philosophy, much aligned to the rest of the team.

Duman Uvatayev – Chief Data Officer
Chief Data Officer since 2012
He earned his undergraduate degree from Central Asian University (2010) and then joined Kaspi’s technology team. Uvatayev leads the company’s data science, analytics and machine-learning initiatives – building the algorithms behind fraud detection, credit scoring and personalized offers. He oversees Kaspi’s “big data” platforms, ensuring customer data from payments, shopping, and finance is analyzed for new product insights
In 2024 Uvatayev co-led a Kaspi delegation (with Pavel Mironov) visiting a national AI research institute, highlighting Kaspi’s interest in applying advanced AI (computer vision, NLP, etc.) to its services, like that of their current Ruslan, a customer support AI, with the CEO claiming it saves 2000 customer facing roles, in an interview.

Management Philosophy:
It is rare nowadays, to find a business which has the same customer obsession, to a similar extent that Amazons’ CEO, Jeff Bezos had, this has been demonstrated by Kaspi repeatedly as below actions:
- Removed internal sales targets and replaced with NetPromoter Score targets, Mikheil in an interview stating that generally, an employee has 1 metric which is tied customer satisfaction which is their individual responsibility.
- Actually sticking to their values, for example Kaspi at the time had a third of their net income from their credit segment (again as per Mikheil’s comments in an interview), however their NetPromoter Score associated with it had always been negative, even after multiple attempts to correct for this, the business ended up killing the credit card segment. Whilst it seems crazy, to me this shows Kaspi is committed.
- Kaspi in 2024 in a single month alone surveyed ~300,000 of their customers, and not through generic surveys either, a large portion of this was manually through emails and detailed calls which the surveying, had been mostly done directly by the development team so that feedback could directly be implemented. This significant amount of surveying has continued.

A large part of Mikheil & the teams’ customer obsession is due to the belief, for a SuperApp ecosystem to be successful, the existing products/services must already be great otherwise the customer will not want to try your new services — Kaspi is able to take advantage of their great services to convert those customers into their new offerings.

Incentives:
As a foreign issuer in the U.S & listed in the local exchange in Kazakhstan, Kaspi isn’t required to provide a granular breakdown, however the following is what I’m aware of:

1. Mikheil, mentioned in an interview, his current ownership being more or less sufficient & would rather use available stock pool to reward staff.

2. Outlined in annual report, LTIP (long term incentive plan) started 2021 so 6.83 million shares will be distributed within 5 years, which is equivalent to 3.6% of the current market cap.

3. Outside of Mikheil, executive directors collectively own 4% still a respectable portion. (as per Kaspi’s own IR presentations)

4. 247 employees (out of total ~10,000 employees) are part of the scheme, seems to match up with Mikheil’s commentary, that compensation is not just primarily going to the CEO in essence rather stock is reserved for directors & key hires. Stock grants are bought for a minimal cost (relative to value). Options vest over five years, in equal yearly chunks, as long as the person stays.

Employee Growth:
As Kaspi has grown, 30%+ (adjusted for currency devaluation) over the last few years, staff count has grown at a slower rate of mid-teens percentage as below:

This is important for the following:
- Shows that Kaspi’s investments in their customer support AI Ruslan & general automation efforts are paying off as Support & Adminstration staff growth has remained near flat, relative to significant growth.

Whilst this is beneficial in terms of saving on cost, important to note, this does not come without drawbacks as staff, as per available sources, such as Glassdoor in their numerous reviews, mentions of significant workload volumes and frequent below market rates, therefore conscious that simply incremental increasing workload is also a driver behind these results, most likely & arguably some risk of some margin compression here, if there workforce marketplace were to increase in competitiveness.
Though I’m sure Kaspi’s brand also has its advantages here.
Additionally I am not expecting this to continue growing operating leverage.

Grey Area / Shady Dealings:
As an investor, I am actively cautious to avoid having a rose-tinted view, here is the breakdown of what you’d need to be aware of, in terms individuals involved within management, whilst a lot of this is arguably smoke, often times, a fire is the cause.

Kairat Satybaldy, nephew of former Kazakh dictator, acquired an initial stake in 2015 (exact investment amount undisclosed & timing of accumulation unclear), reaching 30% ownership by December 2017—coinciding with Lomtadze joining as a shareholder in 2015 with an initial 9.9% stake via a vague “partnership arrangement” with Kim from an unstated earlier period. July-September 2018, Satybaldy dumped his entire 30% ownership on KASE at a time where Kim, a publicly self-declared friend, coincidentally had been buying; six weeks later Kaspi announced the LSE IPO attempt (which was then blocked due to the close timing of exposure of the Politically Exposed Person —Satybaldy, which speculated as the reason behind the rushed sale, it was an attempt to avoid being blocked — though Kaspi successfully listed in London the year after).

Kim spent $390M buying this 30% block of stock in 2018, then transferred the 30% ownership to Lomtadze in December 2018 for “certain non-cash consideration”—undefined mysterious “partnership consideration” later being explained as ‘formalizing partnership after 13 years’. Analysts theorize Satybaldy retained true ownership via a “safe agreement” where Kim/Lomtadze became nominal owners to clean his name pre-IPO, with Satybaldy retaining recall rights on the shares. In 2022, Satybaldy was imprisoned for embezzling ~$500M from Kazakhtelecom & Transport Service Center; he pled guilty to laundering money through multiple Kaspi Bank accounts then purchasing Russian properties with laundered funds—yet, no Kaspi shares were recovered during asset seizures despite $1.5B in assets stripped. Kaspi denied allegations, claiming Satybaldy merely “exited formally,” that Lomtadze “built Kaspi” so “earned” his stake, and that their 13-year unwritten “arrangement” was normal practice due to “complete trust and chemistry”—could be seen as the Kazakh playbook of using shell transactions to obscure politically-connected ownership before international listings.

Though I can’t deny Lomtadze efforts at Kaspi have completely changed the permanent trajectory of Kaspi away from the no-name small regional bank it was only a decade before — so I could also see a rationale argument for this. When considering that as an investor I would personally willingly dilute my hypothetical business by 30% if I knew I could have the same results which Kaspi has had.

However I also note when looking at the lens of Kim, who in previous years, publicly called Satybaldy “a longtime friend and partner” makes me think that I wouldn’t be surprised if the rumours turned out to be true.

However I’m also not as concerned as the former Kazakh inner circle seems to have lost power with insiders such as PM Mamin, Speaker Nigmatulin no longer in office and the jailing of Security Chief Massimov & Nazarbayev being stripped of his Elbasy title in 2022 which gave immunity from prosecution, special privileges, and symbolic authority over the country, power dynamics seem to be replaced by the new party. Additionally what still remains true even if the rumours are true is that Kaspi is a business solely focused on the customer experience and truly operates on these principles, for example Kaspi manually surveyed 300,000 customers in a single month in 2024 & cutting its credit card business (outlined above paragraph).

Whilst there are other red flags, most raised by Culper Research’s Short-seller Report such as:

1. Mysterious death of former lawyer, Derkunskiy, who allegedly made accusations against Kim being a proxy for Satybaldy ownership, more can be read: Article

2. Opaque ownership history, particularly in 2000s-2010s as the Oligarch class was forming & limited disclosure on reason behind share transfers.

3. Kim’s ownership in Magnum Cash & Carry (of which Kaspi owns 90% of Magnum e-Grocery) having been bought from an individual who had been documented to have ties within the inner circle

4. Kim’s alleged daughter Mrs Yulia Kim is running the venture e-Grocery Magnum venture but this was not disclosed in SEC filings, though Kaspi have responded saying she “has no family relationship to Mr. Vyacheslav Kim.’“

Is this great? No.
Is it ideal? No.
Do I think the above behaviour is more prevalent when operating in countries such as Kazakhstan at Kaspi’s scale, particularly during the pre-2019 Nazarbayev era when oligarch influence was stronger, though some influence remains, post-2022 reforms under Tokayev? Loaded question but answer is Yes.
Do I think this makes the whole Kaspi business unethical rather than specific individuals, practices, or institutional lapses? No.

On a Macro level, political changes have improved Kazakhstan’s rating on the Corruption Perceptions Index from a low of 21 in 2007 to 40 in 2024 (Global avg 43), such above exposure may still occur.

Looking at history:
Forms of political involvement occurs everywhere, whilst again, I’m not a proponent of this, it still occurs. On a Micro level when looking back at individual businesses, you can see this is not unique to Kaspi or Kazakhstan as per the below examples::

1. South Korea’s 4 chaebols make up atleast 50% of GDP and hold significant political influence, evidenced in the likes of Samsung’s $38 million bribe in the scandal that ousted President Park Geun-hye.

2. U.S. giants like Google ($12.3 million in 2025 lobbying so far) publicly contribute to political lobbying to influence rules.

3. Lockheed, paid $22 million in bribes across Japan, Netherlands, Italy, and West Germany (1970s), bringing down Japan’s Prime Minister Tanaka and triggering the Foreign Corrupt Practices Act (1977).

4. Goldman Sachs, prioritizing lucrative fees, underwrote $6.5 billion in bonds for Malaysia’s 1MDB fund from 2012-2013 despite red flags of corruption, aiding over $1 Billion in bribes to officials like PM Najib Razak (enabling the embezzlement of $4.5 billion resulting in a $2.9 billion 2020 settlement, executive guilty pleas

5. Boeing Multiple death instances of whistle-blowers, associated with Boeing in essence over ignoring safety concerns and linked FAA probes into falsified records.

6. Ford: 1990s-2000s Explorer-Firestone crisis — 271 deaths and 800+ injuries from tire tread separations, primarily on Explorers. Internal documents showed Ford knew since 1996 about failure patterns in hot climates (Venezuela, Saudi Arabia) but delayed U.S. recalls until 2000.

The above are just a few examples, when assessing the facts the below still stands; Kaspi’s dominant market share exists because their customer obsession and has been clearly proven by their actions, resulting in their offering being that much more superior to anyone else within the country & having generated 40%+ ROIC for over a decade!

Many major companies have historically had operated in the grey, again though certainly not a proponent of the likely bad actors, I’m also viewing reality for what it is & do not want to stray the fact this is an exceptional business which is truly customer focused and has allowed Kaspi to create a moat. The business has a real opportunity to expand multiple times over implementing their playbook internationally whilst paying out high yields — particularly as the business targets 100million user target with it having only had 14million in 2024 (pre-acquisition announcement).

Shareholders and Liquidity

Ownership in the majority is controlled by the following 3 largest shareholders:

As per fillings on the KASE exchange, with no other major institutional ownership from any of the common suspects, which to me shows international institutional exposure is underdeveloped here. The top 3 holders own 68.26% of the business, so obviously between the three entities have control over the company and something to be aware of.

There are only two other institutions with holdings above 1% ownership which are: BAILLIE GIFFORD & CO and FIL LTD who own 2.52% & 1.98% respectively, though immaterial.

TOP 3:
- Baring Vostock, private equity firm which bought into the business in 2006 with it being Kaspiyskiy Bank, then having increased its stake, in the peak to ~38.59% in 2016 and has held most of its share in the business as it has completely evolved (selling some at IPO/exchange offerings). Ownership is structured through it various investment fund structures. Whilst there is a risk of the PE firm selling once a certain time horizon has passed, I see this unlikely for the foreseeable given the support the firm has helped Kaspi with its more recent acquisition by selling part ownership of Kolesa to Kaspi, which is a Car & Real Estate classifieds platform in 2023.

- Mikheil Lomtadze, ex-partner of Baring Vostock, introduced to Kaspi. Mikheil is CEO of Kaspi & would be justified in being called a founder as he lead the business to completely evolve and installed the culture, ruthlessly focusing on customer experience. Mikheil is set on growing the business to atleast the 100million user mark as mentioned throughout multiple interviews/earnings calls, signalling he is here to stay long term.

-Vyacheslav Kim, Chairman of Kaspi, who as mentioned in the intro, had initially acquired Kaspi for the purpose of integrating, financing options into his electronic retail store business, later selling 51% to Baring Vostock.

Whilst there is a risk for one of the major shareholders to completely exit, I would see this as unlikely considering all 3 have held their ownership for just under 2 Decades if not longer and have already taken decently sized paydays on portions of their ownership. Considering the vision Kaspi has to expand internationally I do not see for it to be likely for a complete exit of the position.

In summary the top 3 major shareholders who control the company are the same people/group of people who built the business up into what it is today and have held for almost 2 decades, Mikheil, clearly has helped with day to day operations/long term vision. Vyacheslav has aided with his retail experience which certainly carried over into marketplace offering & Baring Vostock have helped bring Kolesa deal, provided early capital & brought on Mikheil, who has been a world-class operator of this business.

I would still expect them to continue gradually clear some of their positions opportunistically, like that of during listing onto the exchange, in general I would not see it to be a likely risk that one of these holders dump their entire position for irrational short-term reasons. The main thing for me here is I do not see the risk of an institution dumping their complete holding for irrational reasons — e.g ESG, Remit on which countries able to invest in etc - which could cause forced selling.

Liquidity: Kaspi trades around ~600,000 shares per day, at todays current $77/share, amounts to $46.2million/day, somewhat on lower end for a $15bn market cap company, though expected when the top 5 owners account for 70%+ of the entire business but again nowhere near any sort of issues potentially encountered in micro/small cap land.

Growth Runway

Recent Growth

The above image shows performance of the 3 key profit drivers for Kaspi, totalling about a third each. Growth targeted at 20 to 25% for marketplace for the year initially and had come in below expectations at an average growth of +17%, though there are some real one-time headwinds such as regulations around phone sales causing an estimated ~7% slowdown on marketplace I am also not surprised given the physical constraints Kaspi currently operates within. Therefore I take an approach of looking at the factors surrounding Kaspi and assess what growth levers it has to pull.

One of the main historic growth levers (mentioned in the above segment) have been acquisitions, a large part of growth historically and has enabled, to start entering new markets at a faster pace (rather than having to build ground up / spend multiple years going through regulatory hurdles to receive a banking license, in the case of Rabobank A,S).

If Kaspi, is able to continue building and acquiring to integrate businesses into the ecosystem in Türkiye and other future regions, then the long term 100million user target, set out however this is riding upon a big assumption that Kaspi; will be able to successfully integrate the ecosystem in a different market.

Growth however here is not guaranteed internationally therefore I will outline two scenarios:
1. Assuming international growth hinders/doesn’t take off, where do I see Kazakhstan growth continuing
2. Assuming international growth has similar streaks of success to that of in Kazakhstan by following the same playbook.

1. Growth in Kazakhstan:
Longer term, I see this segment sustaining 10% to 20%+ (when normalising for currency devaluation) growth for multiple years out, whilst this isn’t an exact science this is more based on assessing the levers the business has to grow:

1. Increase in usage of multiple features - Kaspi has seen a continual uptick in this segment and not something I see slowing down as there are a number of newer features which had only been added within the last few years.

2. Though similar to the above point, the E-grocery segment in itself particularly has room to grow, as currently the service is limited to 5 dark stores in a few major cities, as the number of dark stores continues expanding particularly in other cities, there will be a sizeable runway ahead on this segment.

3. Incremental increases in AOV over time (normalising to exclude E-grocery), there has been a consistent uptick in AOV, do not see stopping particularly as the per-capita figures are increasing for Kazaks.

4. Bolt on acquisitions & new services added over time such as the Airline/Train ticket sales platform. Management has had to limit opportunities for new ideas previously due to competing priorities meaning there will be a pipeline to continue adding features previously proposed.

However it is important to note that there are limitations in terms of market share gains to be had considering Kaspi had 14million users in Kazakhstan in a country of only ~20million (including children, who obviously wouldn’t be users). Kaspi is aware of this constraint hence their expansion over to international markets.

2. Internationally Opportunity:
It is important to preface that while successful expansion isn’t guaranteed & not enough of a track record in the way of international expansion has been built to clearly prove Kaspi will be able to succeed, I still see Kaspi being a successor for the following:

1. Showed early signs of success in Ukraine investment with 40% growth in year 1 & had initial regulator approval to purchase a bank to then integrate; prior to conflict — plans put on hold.

2. Customer Obsession which Kaspi has, is unmatched against other fintech rivals, paired with a Payments, Banking, Marketplace ecosystem, an already proven successful formula in Kazakhstan, is being applied internationally.

3. Mikheil who ran the business had worked in Private Equity previously, and has the skillset of acquiring businesses unlike other management teams who are often dependant on investment banks. Additionally PE firm Baring Vostok essentially would be one of their advisors, as a major shareholder. Expertise can be seen with the strong hit rate, documented above in Acquisitions.

4. Generally Kaspi is already acquiring successful businesses in the country, therefore eliminating start-up risk & risk of being ‘out of touch’ with the local market, as shown with the Hepsiburada acquisition the business keeps & utilises local management team, and their knowledge of the region.

5. Huge consumer dataset, helps them better understand what works & doesn’t, though this is based on Kazakhstan users, but expect principles to carry over.

6. Arguably similar points could have been made on; Kaspi failing to compete in alternative industries such as: Online Ticket Sales, e-Grocery etc, which have been successful, I believe in part to do with Mikheil’s approach into immersing himself completely into a new area has helped historically.

Should this be successful, Kaspi would open up its TAM significantly, which is important for a business which has already taken a significant amount of its pre-existing TAM. Obviously the growth rate, would be difficult to see; however if this were to workout I would see total growth re-acceleration, with 25% and upwards being realistic for multiple years ahead as, no longer limited by Kazakhstan & long runway ahead, particularly in countries with multiple times the population (Türkiye 86million people). This would bring the business closer to managements ambitious 100million user business, which it has raised multiple times to investor’s and is seriously pursing

Kaspi, carries additional upside when Russia-Ukraine conflict clears up and Kaspi goes to un-pause on growth in its Ukraine operations.

Risks

Here’s the catch; there is certainly hair on this investment idea, all I’m doing in this segment is providing both perspectives.

Monopoly Risk
Risk: Government intervention leads to the business being split up.
I see this unlikely to occur considering Kazakhstan’s Agency for Protection and Development of Competition (APDC) conducted 97 investigations in 2024 (reported Jan 2025) and more still ongoing (including in the banking sector such as identifying signs of unfair competition in Eurasian Bank JSC but none on Kaspi. Currently the ongoing de-oligarchization campaign since 2022 aims to dismantle oligarchic influence through targeting: media control & economic monopolies.

Kaspi is viewed as a “national pride” for digital innovation, with government partnerships e.g., GovTech integrations for taxes/fines — which are in the governments interest to support, as the convenience in their GovTech services, act as a carrot, which has increased tax collections & Kaspi having been a key driver for digital payments hit 95% of non-cash transactions in H1 2025.

The governments’ support is evident, as when money laundering claims came in, they were personally shutdown by the government too. Additionally even if Kaspi were to get split up, which I don’t see occurring, from a sum of the parts valuation, at current valuations, Investors would be unlikely to lose.

Bank Run
Touched on, earlier in the Fintech segment, in 2022, what seems to be the peak of Russia-Ukraine conflict, had triggered deposit withdrawal concerns in Kazakhstan: Sector outflows of 8-12% whilst Kaspi only saw a 4% dip in Q1 2022, mitigated by app alerts and incentives; full recovery by Q2, with 15% year-end growth, this proving the ecosystem-lock in is a real advantage relative to competitors in terms of retaining deposits.

However looking at the worst case scenario, even if mass-scaled withdrawals occurred simultaneously, Kaspi would still be well positioned to survive when considering:
- Business operations are diversified, banking only making up 1/3 of net income.
- Approximately half of loans are less than 12 month duration, more detailed on Fintech segment.
- Balance sheet has always been managed cautiously so it is not over levered.

This to me places Kaspi in a position of strength if any Bank-run scenario were to occur, even more so now, as Kaspi begins to diversify into a new economy (Türkiye, more above in: Acquisitions). Realistically if something like were to occur it would only place the business in a stronger position relative to competitors, after the event.

Currency Risk:
There are 2x main components of this:
-Debts in other currencies, e.gUSD
-Operating currency devaluation

Debt in other currencies: Whilst Debt has been issued of $650mil USD for an acquisition, this is fairly small considering profitability of Kaspi, hence the reason behind cutting the dividend until this is paid off, which is set to be reinstated within the next few months, implying the debt is soon paid, no other issues considering the remainder of their debt is essentially from depositors.

Kazak Tenge Devaluation: Devaluation is real, on average currency has devalued ~9% per year, over the last decade, whilst this is still ok in my opinion, when considering Kaspi is able to achieve 15-20%+ growth on top of devaluation.

I wouldn’t see currency devaluation within future has been devalued further than normal when considering the Tenge has indirectly hit alongside the fall of the Russian Ruble, due to trade ties. Though this is only one of few likely reasons including Oil dependancy, Inflation etc impacting the currency.

Oil Dependant Economy
A large portion of Kazakhstan’s economy is dependant on oil demand, particularly as WTI has been trading at $60/barrel for some time now, there is possibility any further slowdowns will adversely impact growth rates and even trigger weakness in the economy, which Kaspi, being such a large portion of the economy will also get hit by this.

This has possibility to cause a slowdown in Kaspi’s numbers when looking at the short term — though longer term I see Kaspi’s well insulated business fundamentals not being impacted in a material way.

Russia Implication
Kazakhstan is known to trade frequently with Russia on the likes of commodities, this may mean an increased sanctions risk, Culper short report alleges Kaspi of taking on on User reports whilst I do not see this to be accurate considering there are protocols setup in place for example, foreign users are required to go in branch for verification etc

Summary:
Whilst there are more risks here largely in part due to where the business is situated, the business itself has an enduring MOAT which will help the business survive, long after any of its peers, from my experience the management team has generally been prudent in their approach historic, likely due to the business being so starved for cash throughout its history.

Additionally even if some of the risk where to play out, with a forward P/E of 5.76x, personally I still see limited downside.

Valuation

I will always caveat this segment by stating my own opinion on forecasting valuations;

The risk / reward should be so obvious that calculating the exact numbers shouldn’t be relevant, if it fails to be obvious, it’s a Pass.
The below presents simple Napkin maths based on a couple different scenarios:

Base/Bear Case: Growth slows down nationally and international growth is limited, in this situation as mentioned in a prior segment, project 10-15% YoY growth (compared to current 20%+ growth).

Market Cap: $13.94 Billion
YE 2025 estimated Net Income: $2.3 Bn USD
14x Price/Earnings Multiple
12% Midpoint, YoY Growth

YE 2026 Valuation: 12% increase x $2.3 Billion N.I x 14 P/E = 36.06 Billion
Upside: +158.7%

Feel this is sufficient multiple applied given modest growth 10 - 15%, difficulty in terms of future expansion opportunity within Kazakhstan, even factoring in Kazakhstan market discount, adjusting for ultra-high margin diversified business with a True MOAT premium.

Bull Case: International growth successful as ecosystem model starts working, project 25-30% YoY growth for the foreseeable as growth runway has increased significantly.

Market Cap: $13.94 Billion
YE 2025 estimated Net Income: $2.3 Bn USD
23x Price/Earnings Multiple
27% Midpoint, YoY Growth

YE 2026 Valuation: 27% increase x $2.3 Billion N.I x 23 P/E = 67.18 Billion
Upside: +381.8%

Rationale behind applying 23x P/E is: TAM increases significantly as ecosystem model is proven internationally, more diversified geographically, sustained runway for increased growth; alongside earlier mentioned factors.
Overall I have been more towards conservative when using calculating future figures, given uncertainty involved.

What if the stock stays flat?
Not factored into the above is that, Kaspi has confirmed it will start reinstating capital returns to investors, once Eurobond mostly paid (management confirmed over next few months on earnings call).

Whilst this is more of a guide, as per their public offering prospectus & assessing what management has historically followed is with their capital returns is to: Return approximately 50% of free capital back to investors, which at current forward earnings works out to an annual dividend yield of 8.69%.

This alone, assuming stock remains flat, despite trading at f P/E of 5.75x & currently growing 15%+, on dividends alone would match/beat S&P500 average historic returns!