I've Always Been Paranoid About Size

Because I recently republished my first ever Substack, about Value Investing and Art Collections, I’m below including one from May 2024 that I considered Part Two of that discussion. That second post has only had less than 400 views, maybe because the title “Position Sizing for Value Managers” sounded dull, so I livened it up, as many Substackers do (I often take the bait).¹

Part One introduced the concept of “Cézanne stocks,” high quality companies that, unless trimmed or, God forbid, sold, can appreciate for decades, making the whole portfolio’s return. The value of the top five paintings in Keynes’ collection grew over 100 years to 75% of its entire value, with Cézanne’s “Seven Apples” alone being 25%.

In Part Two, I applied the Cézanne principle to fund management. I asked what would’ve happened if I’d frozen my own fund’s portfolio in 1995, leaving our one Cézanne stock and one Matisse stock untouched to grow while my partner and I took a 30-year vacation. Not to give away the answer, but probably Firebird Fund would’ve done just as well, but we wouldn’t have survived to find out. (Past performance is not indicative of future performance etc. etc.)

Most market sages recommend prudent diversification: Bernard Baruch advised selling a position to the “sleeping point,” where you don’t have to worry at night. Joel Greenblatt — who’s a member of my golf club but whom I’ve never gotten to play with, maybe because he heard I’m bad — modified that, saying he’d let a great stock be a large weighting if he believed there was an extremely low risk of a permanent loss of capital.

Exhibiting a bolder attitude, Charlie Munger once said, “When you know you have an edge, you should bet heavily. They don’t teach most people that in business school. It’s insane.” Of course, permanent capital allows Berkshire Hathaway to bet “heavily” without fear of looking greedy, reckless, or just plain bananas, and thereby risking redemptions as fund managers do. That said, Warren Buffett too finally capped the size of Berkshire’s Apple holding, as discussed below.

“Bananas”

Update on the Magnificent Seven’s results since they were mentioned in the May 2024 post: obviously they’ve continued to do very well as a group and, as described in the post, the leading performer has continued changing hands along the way.

Now, here is the original, not boring but boringly titled post.

---

In the first part of this analysis, I wrote about the dilemma an investor in stocks faces when one star performer in their portfolio becomes an outsized weighting. It seems prudent to rebalance the portfolio by trimming the big winner and reallocating to the laggards or using the proceeds to fund essential needs: taxes, champagne, etc. I noted my own experience of buying Facebook at $18 and then cutting the position when I judged I’d “made enough,” then sadly watching it triple again.

To illustrate the attractiveness of a static, non-rebalanced portfolio, I discussed the art collection of John Maynard Keynes, which in 1939 was bequeathed to Cambridge University and has been kept intact since. It turns out that in 2019, 101 years after the collection was assembled, the five most valuable paintings represented 75% of the collection’s value, with one work alone being 25%. Keynes and Cambridge never “trimmed the Cézanne position” and “added to” the other paintings, most of which didn’t became valuable at all.

The Keynes story got me thinking about Firebird Fund, a Russia-focused vehicle that I co-founded 30 years ago this month and that I still co-manage. After reading about the Keynes collection in 2021, I wondered what would have happened if Firebird had kept our fund’s original portfolio intact. It couldn’t be that we would have done as well if instead of all the hard work, all the research and trading, we had gone on a three-decade vacation?! (Caveat: this was before the Russo-Ukraine War and the freezing of Moscow Exchange trading for foreigners; the same analysis done today would raise very different questions.)

Firebird 1995: Including a Cézanne and a Matisse

The oldest Firebird Fund portfolio I could locate was from 1995 (this was pre-Internet, so I had to go searching through paper files). I found that while the fund contained numerous stocks that went nowhere, such as utilities that merged into holdings on unfavorable terms, defense companies that got renationalized, and victims of 1990s-era Russian robber capitalism, it also contained Lukoil and Norilsk Nickel. Lukoil and Norilsk remained Russian blue chips, producing extraordinary total returns over the 26-year period 1995-21. Bloomberg data goes back to the 1990s only on Lukoil, but a comparison of its total return in dollars to Firebird Fund’s net return from 3/31/98-12/31/21 shows that they are almost identical.²

We can assume, based on Norilsk’s total return since 2007 (when Bloomberg data starts), that it performed as well as Lukoil from 1995-2021, if not better. If you have read the first part of this analysis, you know that over a 25+-year period, the return on a diversified portfolio converges with the return on the top quintile. Therefore, with both Lukoil and Norilsk doing what they did, we can conclude that Firebird’s managers could have taken that long vacation, left the 1995 portfolio untouched, and enjoyed about the same performance.

But there’s a catch: had we tried to do that, we may never have survived to enjoy all the ultimate gains on Lukoil and Norilsk. Between 1995-2021, the period under examination, Russian stocks experienced two major crashes (1998, 2008) and three large corrections (2000, 2011, 2014). Firebird Fund entered all of these with reasonably diversified portfolios (in 1998 and 2014, partially hedged with derivatives and cash); though we were down, the vast majority of our investors rode out the turbulence. We communicated frequently with them and what they saw was a stable management team, calmly assessing events and determined to recover the fund’s recent losses. They may have been disappointed with us but had no reason to doubt our sanity.

If, on the other hand, we had entered any of these crashes or corrections with a 75% weighting to just two stocks, Lukoil and Norilsk — which likely would’ve been the case had we kept the 1995 portfolio static — we may have looked like excessive risk-takers to our limited partners. Norilsk fell as much as 70% in dollar terms in the 2007-08 crisis; here’s poor me in this hypothetical scenario explaining to a furious fund of funds manager (correctly, but he doesn’t care) that there’s “no reason” for our 35% position to be down so much. There’s a good chance that many of our investors would have withdrawn from the fund, setting off a cascade of redemptions at the worst time and possibly ending in a winddown.

Concentration and Fund Risk

There are numerous examples of funds whose over-concentration backfired, causing investors to question the managers’ judgment and pull out. The highest profile recent case was Sequoia Fund’s experience with Valeant Pharmaceuticals, a stock that skyrocketed, growing to over 30% of the fund’s NAV, before crashing amidst a scandal. Despite Sequoia’s excellent 47-year track record, its reputation was badly damaged by the Valeant episode, many investors exited, and the manager in charge of the position retired.³

Fund manager Bill Ackman was also punished for Valeant, which he had initiated at 20% of his portfolio: when the position blew up, he faced a 2/3 redemption.⁴ He has since switched to mostly permanent capital, allowing him to take big risks without outflow pressure from investors who may perceive him to have been reckless if the trade doesn’t work out.

The risk to a fund and a manager from over-concentration is not only a perception risk but also a psychological one, as at some level the manager’s objectivity can become compromised. He or she may “fall in love” with the stock and dismiss obvious and growing risks, as for example when Sequoia bought more Valeant even as the scandal was unfolding. While many of Sequoia’s investors likely forgave the original overexposure, they couldn’t get past the doubling down.

The greatest equity investor of our time, who is nothing if not disciplined, has just trimmed his large position in Apple. While Warren Buffett gave no reason besides tax planning, over-concentration likely was part of his thinking. Let’s posit that if Apple at 40% of Berkshire Hathaway’s equity portfolio with its permanent capital structure is too big, then Valeant at 30% of open-ended Sequoia Fund was also too big and go from there.

A Concentration Model Built for Survival

While many value investors run concentrated portfolios, believing that no one can identify more than a few great opportunities at a time, they must also caution against betting too heavily on one or two stocks — especially if they run an open-ended vehicle.⁵ A more manageable portfolio, and one more likely to survive a major correction, would resemble the Magnificent Seven U.S. tech stocks, where one or another position can race ahead (as Apple or Alphabet did, and later Nvidia), but eventually the others catch up — if their financial results deliver. A manager can then tolerate an overweighting in the confidence that it’s temporary, averting the need to keep selling one of their fund’s best stocks — trimming the Cézanne — just because it has grown big (but not 30%).

This is what Firebird has tried to do in Eastern Europe, with 50-75% of our portfolio generally attributable to fewer than ten stocks at any given time. Our belief that all the components of our own Magnificent Seven, Extraordinary Eight, Nice Nine, or Titillating Ten will keep pace is underpinned by our confidence in their earnings fundamentals and dividends. Stock performance can be slow or fast, that’s up to the market, but it should come as long as the positive fundamental trends continue. And in the meantime, we are — and appear to our investors to be — sane and objective, although admittedly it’s hard not to fall in love sometimes.

1. I considered the even more eye-catching title, “She Couldn’t Handle My Size,” but that seemed too much and too ludicrously fictional.

2. Actually, we matched the performance net of fees and expenses, including performance fees, and with diversification and reduced volatility — so arguably a better investor experience. We’ll put that aside for purposes of this discussion. Past performance is not indicative of future performance.

3. E.g., Sequoia Fund: Valeant Bruised Our ‘Credibility as Investors’, WSJ, February 29, 2016

4. Ackman Abandoned by Big Pershing Square Investors, Financial Advisor, April 5, 2018

5. Not all value managers believe that concentration is necessary. Our own U.S. value strategy, for example, targets a relatively equal-weighted portfolio of 30 stocks.