Nassim Nicholas Taleb, a celebrated expert on complex risks, in a video Taleb has graciously allowed Naked Capitalism to run exclusively, challenges CalPERS Ben Meng’s rationalizations of his decision to dump both of CalPERS’ two “left tail risk” hedging strategies. That call resulted in CalPERS giving up over $1 billion gain, or a 3600% return, on the bigger of the two hedges that CalPERS closed out first, prior to the market free fall.
As we’ll show below, Taleb is reacting to Meng’s self-defense in a webcast on Wednesday. In addition to questioning Meng’s understanding of these portfolio hedges, Taleb charges Meng with misrepresenting the supposed hedges that Meng said CalPERS had in addition to the two tail hedges CalPERS abandoned. Meng asserts that CalPERS made $11 billion from these positions. Taleb contends that those investments could hardly be considered hedges; his quick and dirty math says that in 2019, they lost about $30 billion, still leaving CalPERS $19 billion net worse off than it it had left well enough alone.
How can Taleb know that? First, in his talk, Meng specifically mentioned his two “risk mitigation” strategies, which were “long Treasuries” and factor-weighted investing for equities. One might be able to infer a bit from CalPERS’ Investment Committee meetings, but Taleb has a much better fix than that.
Taleb is an adviser to Universa, the manager of the position that CalPERS exited first. A senior investment manager at CalPERS, Ronald Lagnado, who had been head of asset allocation at CalPERS, joined Universa at the beginning of April, straight from CalPERS. In other words, Lagnardo is ideally positioned to the details of how these strategies were implemented, which would enable Taleb to estimate their historical performance.1
It’s not as if Universa and Taleb need more press. In addition to the CalPERS’ self inflicted wound calling attention to Universa’s sparkling performance in the equity plunge, the Wall Street Journal wrote up Univera’s great results on April 8. By then, its gains were even higher: 4144% year to date. And the Journal effectively debunked Meng’s assertions that his other secret sauce risk mitigation strategies were cheaper substitutes for the supposedly prices Universa hedges. In fact, classical diversification strategies, which is all Meng appears to have to fall back on, barely dented the damage:
Mark Spitznagel could be forgiven some immodesty. His Universa Investments, which offers investors a tail-risk hedging strategy that serves as an insurance policy against extreme market events, made a return as crazy as the market action this year: 4,144% in the first quarter.
That is the net return of the strategy on its own, but Universa expresses it as a small part of an overall portfolio. For example, last month the S&P 500 lost 12.4% of its value while an investor with 3.3% in Universa’s tail-risk hedge strategy and the remainder in an index fund tracking that stock market benchmark would have made 0.4%. No other “risk mitigation” trade, such as putting a chunk of a portfolio in gold, bonds or a basket of hedge funds, would have had a positive return.
Yet a fantastic month for stocks wouldn’t have meant the inverse because Universa doesn’t make linear bets—its frequent losses are small.
In other words, Taleb appears to be pursuing his mission of calling out people like Meng, who wield power in the investment business but don’t know what they are doing. The video is short so please watch it in full. Nevertheless, I’ll call out the key points below. I’ve also posted the CalPERS webcast that Taleb mentions later in this post.
As you can see, Taleb presents the purpose of his short talk as clearing up misconceptions promulgated by Meng about tail risk hedging. He first addresses Meng’s assertion, which he has made in the press as well as in his presentation below, of having “alternative hedges” that were cheaper and by implication largely as good as the Universa hedge and the smaller exposure that CalPERS had with a second tail-risk hedge provider, LongTail Alpha. For instance, from Meng via Institutional Investor:
We terminated explicit tail-risk hedging options strategies because of their high cost, lack of scalability, and the fact that there are better alternatives available to CalPERS.
And from Meng in the video below, starting at 10:47:
In December 2017, our board approved risk segmentation as drawdown mitigation managers as cheaper and more scalable drawdown risk mitigation than explicit options-based tail risk hedging strategies. As we continue to enhance our balance sheet liquidity management and develop a centralized governance structure, as part of the broader active risk review in 2019, the termination of an expensive and unscalable strategy became apparent.
Let us stop here, since there are problems with what Meng says in addition to the whooper Taleb called out. Meng attempts to depict the tail risk hedges as intended for liquidity risk mitigation, as in to make sure the fund has enough cash on hand even in bad markets so as to be able to make required payout without selling assets that might be distressed with the time. The conundrum is that highly liquid and low risk investments don’t return much, so investors like CalPERS endeavor to provide for the needed foul-weather liquidity while minimizing the loss in investment returns.
That was not the reason for entering into these tail hedges. Former Chief Investment Officer Ted Eliopoulos entered into these programs to reduce losses in a crisis. Even though loss mitigation and liquidity risk mitigation are overlapping concerns, they are distinct, as you can also see from the Pensions & Investments article describing Ronald Legado’s new role at Universa (emphasis ours):
Mr. Lagnado will initiate research about tail-risk hedging strategies across asset classes as well as drawdown management strategies that help pension fund staffs manage liabilities and position their portfolios for long-term investment.
Translation: Universa sees tail-risk hedging and drawdown mitigation as separate products. But Meng would have you think they are the same thing.
One finally wonders how Meng conducted “active risk review” was conducted. Wilshire Associates, CalPERS’ main investment consultant, was kept out of the loop. They were told CalPERS was abandoning the tail risk strategies after the decision had been made, and even then only in passing.
Back to Meng at 11:20:
So, basically we chose better alternatives for market drawdown protection and they turned out to be better alternatives in the recent market rout. Those two risk segments added more than $11 billion of our drawdown mitigation matrix.
Taleb called out this bafflegab (at 1:07):
Number one, Mr. Meng said he made $11 billion on alternative strategies that sort of offset the losses during this collapse.
I don’t know if you realize that these strategies need to be weighted against what they made or lost before that. Effectively, we think, back of the envelope calculation is so-called mitigating strategy would have lost something like $30 billion the previous year. So you make $11 billion, you lose $30 billion before, not a great trade, and definitely not a great trade if you take that over long periods of time, where you lose in rallies and make back a little bit in the selloff. That’s not a mitigating strategy, that’s something that may work in the portfolio, definitely not comparable to tail hedging.
Taleb’s second point is “all wine is not the same,” or as he puts it, “There are tail risk hedges and there are other tail risk hedges”. Taleb stresses that options can be pricey or cheap, it takes real analytical skill to tell the difference, and the combined 34 year track record, first Taleb’s, then Universa implementing the same type of strategy, shows that they can construct these hedges in a way that is profitable over time.
Basically, Taleb is arguing that options are not always efficiently priced, which enables him and Universa to deliver alpha. His results support his claim.
Wilshire Associates clearly agreed. Recall that Wilshire spoke approvingly of the tail risk hedging strategy at an August 2019 board meeting, saying that even though the month-in, month-out burn might look worrisome, when things went sour, the return would be on the order of 1000%. As we now know, it was four times higher. So Universa spectacularly outperformed Wilshire’s expectations, yet Meng is still trying to portray overriding Wilshire as a sound call.
Finally, one comment, from Meng, which to me is the worst. Meng said that a certain gentleman by the name of Ilmanen, Antti Ilmanen, who used to work I think for Cliff Asness at some point, some kind of economist, said that options aren’t cheap. I’m going to say one thing, I didn’t even mention his name in my book, I know his argument, that options are expensive and one should sell them. He just doesn’t know probability. You have to start by teaching them probability, it will take maybe five to ten years, and then explain to the point. Some people, you can’t deal with their argument.
Taleb charges both Ilmanen and Meng as having failed to train themselves adequately on the nature of probability to opine on options pricing. And there’s a famous example that refutes Ilmanen’s argument that you should only sell options and never buy them.
Credit default swaps, although misnamed as derivatives (they are not priced in relationship to an underlying instrument) have options-like payouts. Those who sold CDS on BBB subprime risk from 2005 to 2008 either went bust, like AIG and the monolines, or needed bailouts in order to survive. The buyers of those CDS (either directly or via CDO tranches), like Magnetar and John Paulson, made fortunes.
Meng’s rambling talk below is hard to digest. Nevertheless, a few points stand out.
First, Meng insists that even with the benefit of 20/20 hindsight, what he did was correct. That’s quite a statement, when you contrast the stellar performance of the tail hedges (again, with Wilshire confirming the eventual payoff would be spectacular) versus the vastly high cost of the risk mitigation strategies that Meng falsely depicted as cheaper.
Second, Meng cited two “papers” to justify rejecting tail risk hedges. One was the Antti Ilmanen article that Taleb could not bring himself to treat as serious: Do Financial Markets Reward Buying or Selling Insurance and Lottery Tickets? We’ll defer to Taleb’s view that the piece is too ill-informed to bother parsing.
The second, Who Should Hedge Tail Risk? by Robert Litterman, is not an analytical piece but a long editorial. The abstract in the Financial Analysts Journal reads: “The executive editor discusses his views on an issue of interest.” 2
Mind you, we’ve gotten this far without reminding readers of Meng’s biggest sin, of lying to the board about exiting the tail hedges. Recall that board member Margaret Brown asked him a direct question at the March board meeting and Meng gave a specific answer. From the transcript:
BOARD MEMBER MARGARET BROWN: Ben, can you tell me how our left-tail investments are performing? Are they performing the way we thought they would in this economic downturn?
CHIEF INVESTMENT OFFICER MENG: Good morning, Ms. Brown. Yes, for any left-tail risk hedging strategy you’re referring to, they should perform well in this kind of a down market, as they were exactly designed to do. And from what we know are most of these strategies are performing as anticipated.
Under California law, the CalPERS board has sole and exclusive fiduciary duty for the assets of the fund. Fiduciaries cannot afford to have a liar in charge of the money.
Meng is too dishonest, top to bottom, to continue in an executive role. He need to go.
1 Meng refers to two strategies in his talk. One the factor-weighted equities, CalPERS presumably pursued a low-volatility approach, which would mean leaving money on the table in a bull market. On “long dated” Treasuries. CalPERS holds positions in the 10 to 20 year range. In addition to swings in value in this maturity range, recall also that CalPERS implemented a policy of levering its portfolio, as opposed to levering within specific asset classes. What this means in practice is levering where it is cheapest to buy more assets. One of the cheapest ways to borrow is repoing Treasuries. As the saying goes, leverage amplifies losses and gains. If CalPERS opted for longer repo tenors, that would increase the magnitude of the swings. Knowledgeable readers are encouraged to weigh in.
2 The piece is deeply flawed. It depicts investment banks as logical buyers of tail risk insurance, depicting them as levered traders of equities. Huh? How could he have worked at Goldman and not understood their economics? Institutional equity firms trade shares on exchanges. They may do various sorts of basket trades and run some short-term risk in laying them off successfully. The way they lose money in equity downdrafts is that their institutional sales business is unprofitable absent IPO and secondary offerings, and those dry up in bad markets. They are leveraged traders in bonds, and anything but the safest bonds swoon in market crises, but that was not what Litterman was talking about.
Similarly, it gives a naive view of pension funds as long term investors. In fact, CalPERS has to make payments month in and month out, and is now in a negative cash position, which means over time, it is spending more than it is getting in via new contributions.