Introduction
Berczy Park Capital (BPC) is a single-family office formed in August 2012. Our focus is absolute returns, not relative returns. We use a 10% hurdle rate based on historical returns for equities as an asset class and because it allows for decent compounding. Our definition of intrinsic value (IV) is a price range from which a 10% CAGR is likely. The investing style is probabilistic, so there is an expectation that at least a third of our positions will be duds, and that prediction has been correct so far!
Before going any further, we asked ChatGPT to write a disclaimer for this Substack:
Disclaimer: Not Investment Advice
The content on this Substack is for informational and educational purposes only and should not be construed as investment, financial, legal, or tax advice. Berczy Park Capital is a single-family office, and the views expressed here are solely those of the author(s) and do not reflect the opinions of any other entity.
Nothing on this Substack constitutes an offer, solicitation, or recommendation to buy or sell any securities or other financial instruments. Any investments mentioned are discussed solely for illustrative purposes and may not be suitable for any particular investor. Readers should conduct their own research and consult with a qualified financial professional before making any investment decisions.
Berczy Park Capital does not accept any liability for reliance on the information presented. The accuracy, completeness, or timeliness of the information is not guaranteed. Investing involves risk, including the possible loss of principal.
By accessing this Substack, you acknowledge and agree to these terms.
This Substack assumes the reader is sophisticated and has access to equity research, especially Viking’s book and model, which he kindly shares on thecobf.com for free. Viking is a Vancouver based investor who has written extensively about Fairfax on The Corner of Berkshire website. He is considered by many Fairfax investors to be the most knowledgeable investor who covers the company, including all the professional analysts.
FFH has the benefit of institutional research coverage, so our analysis focuses on areas where variability should be expected. We used Koyfin for earnings estimates and other financial data.
The World has Changed
We believe market structure has changed dramatically since the Great Financial Crisis. Most active investors (institutional and retail alike) have gravitated towards stocks that screen well for quality and have been vastly rewarded for doing so. Never sell, quants and passive flows have exacerbated this trend. A side effect is that there seems to be an abundance of idiosyncratic opportunities that don’t screen well but offer exceptional expected returns, potentially well above our relatively meager 10% hurdle rate.
We’re practitioners and not academics, so we appreciate this is a claim that isn’t easy to prove, but in speaking with other probabilistic investors, they also see an abundance of opportunities, while heuristic investors using the value factor lament there are no good opportunities. We believe this dichotomy supports our thesis.
Generational Opportunity
Despite a plethora of idiosyncratic opportunities, Fairfax Financial (FFH) stands out to BPC as a generational opportunity and thus represents ~49% of net assets (~24% at cost) based on our conviction that there is a high probability for ROE to exceed 15% (perhaps meaningfully) for the foreseeable future.
Clearly, 15% is above our hurdle rate, but also above what the market expects FFH to achieve, which is a recipe for multiple expansion. In fact, that’s exactly what has happened in the last four years. FFH averaged ~19% ROE from 2021-2024E, and the P/B multiple has expanded. Given the low relative and absolute valuation, we think it should continue.
Based on a 3:1 investments to equity ratio, improving underwriting, and cheap equity investments, FFH seems likely to double its book value over the next 5 years (ignoring the dividend for ease of calculation). In fact, the P/B multiple could fall to 0.8x over the forecast period and still meet our 10% hurdle. That’s an extremely high margin of safety, especially in the context of excess capital, stock buybacks, and a potential index add leading to price indiscriminate ETF demand.
The P/B multiple was below 0.8x a few short years ago, but that induced FFH to buy back shares and enter into a series of Total Return Swaps (TRS) for 1.96m shares, reducing the net float by ~8m shares since 2017. Buybacks have been consistent throughout, including as recently as last month.
Further, FFH seems likely to enter the S&P/TSX 60 (60) as soon as March 2025. We believe stock prices reflect supply and demand, so adding a new ~4% (of the float) shareholder who ignores the value factor and buys shares almost every week bodes well for multiple expansion. When Intact Financial (IFC) went into the 60 in March 2022, its P/B multiple was ~2.2x. Almost three years later, IFC’s P/B multiple is ~3.0x despite only ~10% growth in BVPS (after dividends) during the period.
On a FTM basis, FFH is expected by analysts to earn IFRS EPS ~$145/sh (down from ~$155 a few weeks ago) which implies it trades at ~10x NTM P/E despite having an expected ROE of ~14%. Most investors, however, use adjusted EPS for their screens, which shows a consensus of ~$127, which is the RBC’s analyst definition of core earnings and ignores all gains/losses in the investment portfolio. For these investors, FFH looks more expensive at ~11x NTM P/E and a ~12% ROE. FFH still looks cheap compared to the market, ignoring gains/losses on the $70b investment portfolio, but it does not screen as compelling. We think how FFH chooses to report (no adjusted EPS) and how analysts choose to respond (mostly all IFRS estimates) impacts how FFH is valued by the market.
Further, we contend that analyst estimates are conservative for predictable reasons. First, the valuation is undemanding, as noted above, so there is no need to stretch to justify a buy rating. Second, FFH earnings streams are historically volatile, although the recent adoption of IFRS17 and FFH’s decision to closely match the duration of the bond portfolio with claims liabilities has reduced the volatility in earnings. Third, after the stock price disappointed investors for a decade (2010-20), they are entirely focused on the left tail, so it’s unsurprising that most of the analysts that serve them are similarly focused. By analyzing P/E on a NTM vs TTM basis, we find that analyst estimates are persistently conservative, and we expect forward ROE to exceed consensus.
Variant Perception
FFH is a complex company, and that’s one of the reasons it’s such a bargain, but we believe there are two key areas for variant perception. First, in underwriting, and second, in income from investments, including capital gains. If we are to distill our fundamental thesis in a run-on sentence, we believe FFH estimates are too low for underwriting and investment income, and that right-tail surprise could result in above-trend ROE for years to come. This is a good time to remind the reader that the recent trend ROE is mid-to-high teens ROE.
Despite the strong recent performance that has a lot to do with very strong float growth and what we consider “normal” interest rates, analysts remain pessimistic about FFH forward returns as demonstrated in the below comp sheet. The margin of safety is high because the stock still looks absolutely cheap on its own and vs peers at only ~10x NTM EPS which meets our bogey but we believe is below the hurdle rates of the few active investors remaining that still use hurdle rates.
Underwriting surprise
FFH targets a 95 combined ratio, so it’s unsurprising that analysts expect the NTM combined ratio to be close to 95. This is a reasonable but potentially conservative assumption. For context, in the first nine months of 2024, FFH has booked a 93.9 combined ratio.
Anecdotally, we hear from many investors concerns that the hard insurance market may be ending, suggesting that combined ratios might be heading higher, and this is a cue for them to sell. We believe these views are represented in analyst estimates. In fact, the recent decline in consensus NTM IFRS EPS seems to be due to expectations for higher catastrophe (CAT) losses in Q424 and Q125.
We believe that analysts are overly focused and likely overestimating recent CAT losses. If analysts are correct, pricing will stay hard, and losses should be recouped over the forecast period, so it’s not a concern for long-term investors. We believe hard markets are very supportive of float growth, as FFH has enjoyed over the 4+ years but a portion of underwriting profit might be delayed depending on a company's approach to reserving.
Based on historical performance and scuttlebutt, we think FFH is conservative when it comes to reserving such that some of the benefits from the hard market will show up in the combined ratio over the next four years via reserve releases. The last two quarters showed an uptick in reserve releases, so we’ll see if that continues in Q4 when a more in-depth actuarial review is undertaken and into 2025.
It’s impossible to predict what future combined ratios will be in any given quarter or year, but over five years, it’s possible that combined ratios could come in as low as 92, helped not only by strong reserving and subsequent reserve releases but also by operating leverage. While analysts understandably predict closer to 95, each point in combined ratio upside (i.e. lower than expected) equates to ~$10/sh in EPS or ~1% in ROE. We don’t need it to happen, but at the current valuation, it’s certainly not priced in.
Underappreciated Upside: Why Fairfax's Investment Returns May Surprise
Consensus estimates are expecting returns of between ~6-6.5% on FFH’s ~$70b investment portfolio. Viking is an outlier above the range, which helps explain why his, in our view, more realistic 2025 EPS estimates are ~7% higher than consensus. We borrowed the below table from Viking’s model, replacing the 2024-2026E estimates with those from RBC as it’s more representative of consensus.
With ~$50b invested in fixed-income securities earning a somewhat predictable ~5% return given portfolio positioning, consensus implies the ~$20b invested in non-fixed income (including the TRS) returns ~9.4%. That sounds reasonable on a macro level, but when looking at the portfolio on an idiosyncratic level, it seems too low, given sizeable contributions from somewhat predictable sources.
FFH owns ~1.2b shares of EUROB, and if earnings are flat on a FTM basis (analysts expect down 10%), it contributes ~2.4% (of 9.4%) on its own. Poseidon, based on guidance laid out in Prem’s 2023 Letter to Shareholders, contributes another ~1.1% given FFH’s 43% ownership. Eurobank and Poseidon are both equity accounted so FFH’s share of earnings flows through to Share of Profit of Associates. FFH is also long TRS (1.96m shares) on FFH stock. Assuming FFH earns a NTM 15% ROE and the P/B multiple increases to 1.5x, the TRS adds another ~5.1%. Those three sources alone get us to ~90% of the expected return but are only 20% of the non-fixed income portfolio assets.
If the remainder of the non-fixed income portfolio outside of the big 3 earns 10%, overall ROE increases by > 4%. We started with consensus ROE ~14% and arguably the true over/under is actually closer to ~20% when using what we consider realistic assumptions on underwriting and idiosyncratic returns in the portfolio into account.
Analysts are reluctant to predict ongoing one-time gains for the non-fixed income portfolio or an improvement in underwriting, which is understandable, but that means the earnings estimates are very likely to be beaten, and that generally helps multiple expansion as quants appreciate earnings beats and price momentum.
Other potential book value boosters
Besides gains like Stelco, Peak Achievement, and a recent forced sale (for Greek regulatory concentration limits) of 80m shares of Eurobank that are either known or somewhat predictable gains, there are other potential catalysts that may occur over the next few years to boost ROE.
Fairfax India
We recently returned from a tour of India organized by FFH holding Thomas Cook (to be clear we paid our own way). We recommend the trip to any FFH or FIH investors and I hope they do it again next year. We met with management teams of a variety of Fairfax India (FIH) portfolio companies. The largest and most pertinent to this discussion being the Bangalore Airport (BIAL). We were thoroughly impressed by all of the management teams, but BIAL was the only asset we got to see in action and represents most of the value of FIH at fair value. It’s the most impressive airport we have ever seen, and it should be noted we are not well traveled but fellow investors on the trip who are more so agreed with our assessment.
FIH owns 30.4% of BIAL directly with the remaining 43.6% via holding company, Anchorage of which FIH owns 88.5%. FIH has disclosed it is committed to best efforts to IPO Anchorage by September 2025. We wouldn’t be surprised if that timeline slips, but with continued FCF growth, it will ultimately mean higher proceeds.
An IPO should lead to an increase in the mark on BIAL. We estimate $400m+ in 2026 EBITDA, which at 20x leaves a ~$6b equity value, which is meaningfully higher than the current ~$2.6b mark. On a look-through basis for FFH, given its 43% interest in FIH, an IPO could boost ROE by another 3%+. FFH would also accrue a performance fee that would boost ROE 1%+.
BPC has a ~5% position in FIH in addition to our position in FFH, and we expect FIH stock to respond well to an IPO announcement. Once Anchorage is listed, most of FIH’s portfolio will be publicly listed, which may help close the BV discount. It’s hard to be confident the discount will close permanently given that FIH is not in any benchmark and thus is subject to the whims of the marginal value investor. However, if Anchorage achieves a fair valuation, management may be able to use valuation arbitrage to sell some of Anchorage and buy back FIH stock, which may keep the discount reasonable.
KI Insurance
FFH has had big winners on the insurance side of the business as well. India-listed GoDigit is a good example of an investment of just over $150m that is now worth over $2.5b. The next big winner we can see on the horizon is Ki Insurance (Ki).
Ki started inside of UK’s Brit, one of FFH’s insurance subsidiaries, as a joint venture with Blackstone but was separated last month to a stand-alone company. Based on our reading of the shareholder agreement we think an IPO is likely in the next 12-24 months. We think an AI-based, Blackstone-backed, profitable fintech and Lloyd’s market insurance company that has grown from nothing to over $1b in premiums in under five years might get a lofty valuation of 5-10x premiums.
FFH deployed its often-used financing method for minority interests in insurance subsidiaries, which include preferred shares with so much protection for the holder (in this case, Blackstone) that they are considered common equity for accounting purposes until they are repurchased/redeemed. While it looks like FFH owns 20% of Ki at a cost of $100m based on the filings, that should change on an IPO to just under 40% depending on pricing as the Class C shares par value will be repaid in common shares at the listing price including a preferred 8% return.
It’s possible, FFH’s initial $100m investment may be worth $2-4b depending on valuation. Another possible grand slam from a start-up insurance company for FFH. The implication is a ~5-10% pre-tax boost to ROE in the year that it happens based on our analysis, which we concede may be optimistic, but once again, it’s not like these catalysts are priced into the stock by any means. As brilliant investor and fellow FFH shareholder Charlie Frischer often says, if it was Berkshire pulling off these wins, it would be a frequent topic of discussion for Buffett acolytes. In contrast, most FFH shareholders don’t talk about Digit except concerns about valuation and don’t know about or don’t appreciate Ki’s potential.
Optionality in Asset Allocation
With almost $30b in premiums, $40b in float, and $50b in fixed-income investments, mostly in treasuries, it stands to reason that FFH could reallocate a significant portion of the fixed-income portfolio to corporate bonds if credit spreads widened meaningfully or into quality equities when there is a broad-based sell-off.
FFH should provide an update on how much excess capital the insurance companies could dividend out without regulatory approval in its 2024 annual report. Presumably, $4b+ could be shifted to equities when an opportunity presents itself. FFH is a rare stock that gives the holder more cash equivalents than every dollar invested and has “buy the dip” built in. We hope that helps cement home why we feel comfortable committing almost half of our capital to FFH.
While waiting for the inevitable market correction, FFH is using its excess capital to buy back stock and buy in minority interests of the insurance subsidiaries that were sold when FFH was growing faster than its capital position allowed. Given how these deals are structured, these investments have very high returns. Most recently, Brit’s minority interest was repurchased at ~5x earnings and <1x BV.
All of these transactions boost future ROE, supporting our view that the current consensus underestimates FFH’s earnings potential.
Short term catalysts
While we are long-term investors, our training on a prop desk helps us recognize short-term opportunities as well (with varying success). In the past 3 years, FFH has outperformed the S&P/TSX Financials ETF (XFN) in the period a week before Q4 results to the end of May quite meaningfully.
This makes sense as book value tends to grow 3-5% per quarter, and the forecast period contains two quarters (Q424 report is expected on Feb 13). Further, the annual shareholder letter is released and AGM takes place during the forecast period and normally encourages value-oriented idiosyncratic investors to add to their positions given new insights gained through incremental disclosures.
In 2025, this outperformance could be supercharged if FFH is added to the S&P/TSX 60 index in March, which may result in demand for ~4% of the float or ~800k shares. Passive and quant are two major sources of flows that ignore the value factor, so they could result in price discovery, i.e., multiple expansion.
Our analysis as displayed in the chart below suggests that Q324 was the first quarter where the 60 Index Committee could have chosen to replace Algonquin (AQN) with FFH as AQN had fallen below 20bps in weight for the first time. This threshold is understood to be a rule of thumb for removal, from the 60. Prior to that, there were bigger alternatives to FFH when a spot opened up in the benchmark. AQN remained below 20bp in Q424 and FFH was once again passed over. While some market participants believe this implies that sector representation matters more than size, we disagree believing instead the odds increase every index review FFH is passed over.
As of Feb 6, FFH has a weighting of 118bp, is 23rd ranked, and the 7th most profitable company in the S&P/TSX Composite (XIC), and we expect FFH to be added to the 60 in March or a quarter soon after. The delay has allowed FFH to continue to buy back its shares with an additional ~327k shares repurchased in Q424 and ~23k shares so far in Q125, so more delays would be welcome.
We don’t intend to sell any of our shares if FFH is added to the 60 as growth and multiple expansion should continue for years, but we believe it’s prudent to be positioned ahead of an event that we think is inevitable. As noted previously, IFC’s P/B multiple increased by ~0.7x to ~2.9x since it entered the 60 in March 2022 despite underwhelming book value growth, with multiple expansion providing most of the returns.
The Very Long Term
We believe most investing has become focused on relative returns instead of absolute returns which has helped lead to a very bifurcated market. Over the very long term, since inception (1985) and over the past 4 years, FFH has compounded BVPS by over 18% per year.
In the medium term (5 years), we expect FFH to achieve a 15-25% ROE range and believe a P/B multiple range of 1.4-2.5x is sustainable. At the low end, that’s a 15% annual return, at the high end it’s 40%+ annual return. Multiple expansion and ROE could always surprise so these numbers may prove conservative.
Our medium-term forecast is based on what we consider the foreseeable future, but we hope to hold our shares until we forecast NTM ROE below 10%. Based on the current setup, it’s hard to predict when that could happen, given the investments to equity leverage and current interest rates, but eventually, it is likely to become part of the foreseeable future.
We are sharing our views primarily for feedback and also so that our fellow shareholders may consider the right tails that FFH has on offer in both BVPS growth and multiple expansion as the stock reaches new all-time highs. We leave you with a reminder to do your own analysis and reread our disclaimer on the first page of this Substack.
Yours truly,
Ryan and Asheef
Thanks for reading Berczy Park Capital’s Substack! Subscribe for free to receive new posts and support my work.