Azam Ahmed’s big article on tail-risk funds got a lot of play in the NYT yesterday, but I don’t know why — he gives us no evidence to believe that they’re large, important, or even growing particularly fast. There aren’t many hard numbers in the piece, but the ones he does supply seem small to me:
Products linked to an index known as the market's "fear gauge" total nearly $2.5 billion. And in the last year, the amount of money managed in dedicated tail-risk accounts by the bond giant Pimco has doubled to $23 billion.
Boaz Weinstein, a former trader at Deutsche Bank who lost more than $1 billion of the bank's money during the financial crisis, began raising money for his own Armageddon fund late last year. It has since grown to $400 million of mostly institutional money, part of the $3.3 billion he has raised for his hedge funds.
The only remotely significant number here is Pimco’s $23 billion in “dedicated tail-risk accounts” — but I’m a bit suspicious of that number, given that Ahmed claims it has doubled in the past year. If you go back to last year’s iteration of this article, from Bloomberg, it was pretty clear that the dedicated accounts hadn’t even launched yet:
The Pimco Tail Risk Hedging Fund 1 will be the first in a potential series of partnerships, according to a private placement filed with the U.S. Securities and Exchange Commission on June 23. The initial fund will be designed to protect investors from a drop of more than 15 percent in a benchmark index that Bhansali declined to identify.
So it looks as though Ahmed is including, here, funds which are designed to go up in value in good times and which simply incorporate a certain amount of tail-risk hedging — funds like the Pimco Global Multi-Asset Fund. That’s emphatically not a dedicated tail-risk account which offers costly insurance against extreme market events.
Ahmed’s article, then, insofar as it purports to be reporting actual news, rests mainly on very vague statements about how “investment professionals have a new pitch,” or what “clients are suddenly realizing,” or that “Wall Street lawyers say money manager clients have approached them in recent months about forming new funds aimed at providing protection.”
And as an analysis of tail-risk funds, it’s weak. There are pro forma to-be-sure grafs about how “protection does not come cheap and occasionally fails to work”, but there’s no indication that tail risk hedging is, conceptually speaking, pretty much impossible. Emanuel Derman gives one reason why: tail risk is not a simple and identifiable thing which can be insured against.
The value of a portfolio can be substantially destroyed by more than one cause. Portfolios can be ruined by equity crashes, credit spread widenings, bond defaults, high interest rates, sustained inflation, increases in volatility, illiquidity, etc. To protect your portfolio against so-called tail risk may require spending money on insurance against all these risks, and there is no panacea here.
David Merkel gives another: if you’re buying insurance, you need to be sure of your insurer’s ability to pay out in the event of a catastrophe. But in the markets, your insurer is precisely the sort of institution which is likely to be going bust as a result of some unexpected tail event. Writes Merkel:
In order for tail risk to be mitigated fairly, someone must keep a supply of slack high quality assets. Rather than the insurer doing that, why not have the investor do so? The insurer brings along his own cost structure. Why not self insure and bring down the risk level directly. Someone has to hold high-quality assets to mitigate risk; let the investor be that party; embracing simplicity and enjoying reduced risk without the possibility of counterparty failure.
This is pretty much the same conclusion I came to last year: the best way to hedge against tail risk is to put 90% of your assets in Treasury bills or other super-safe assets. This is not an extreme strategy at all; indeed, it makes perfect sense for anybody who’s both rich and conservative, including Suze Orman (“I have a million dollars in the stock market, because if I lose a million dollars, I don't personally care.”)
The problem with tail-risk hedging is that everybody doing it wants to get healthy returns during good years and then have very little downside risk during bad years. It’s an impossible combination to achieve, and although Wall Street will happily sell tail-risk hedging products to those who want them, there’s absolutely zero evidence that they actually work in practice.
So if and when Universa comes out with a black swan ETF this month, my advice will be to stay well away. Indeed, insofar as it helps makes investors overconfident, on the grounds that they’re hedged, it’s likely to be positively dangerous — just as portfolio insurance was in 1987. Tail risk, it turns out, is just too amorphous to hedge. Sorry.
I would want to distinguish between an event like the 1987 crash (or LTCM blowing up) as compared to a prolonged bear market like after the tech bubble burst. In a CVaR optimization framework, depending on your preferences, it may make sense to pay a little extra to reduce the amount you could lose in a 1987-type situation (of course, depending on the volatility and shape of the tail, you may pay more or less in different periods). An even like a prolonged bear market, can be better thought of as a separate regime (see Ang & Bakaert in 2002). In this situation you might still want to have some tail risk hedging, but you also would want to significantly reduce equity exposure and increase bond and cash exposures as well.
No such thing as a super safe asset. In times of high inflation you’ll be losing money hand over fist in T-Bills, not to mention rollover costs and assuming you can find ultra-short end government credits.
“Treasury bills”? My brother recently told me that Ron Paul is too status-quo oriented, and what we need is a currency credibly backed by ammunition.
But, yes, anything that does well during good times is suspect. Also, anything that does well during mediocre times. “More money has been lost reaching for yield than at the point of a gun." — Raymond F. DeVoe, Jr.
“the best way to hedge against tail risk is to put 90% of your assets in Treasury bills or other super-safe assets”
That’s not really a hedge, it’s more like putting it under a very secure mattress. Wouldn’t a hedge provide some upside during a down period to offset some of the losses that occur when the unexpected happens?
In property, they have sold “all risks” coverage so I suppose one need not identify any tail risk, just that there is risk. I wonder if the biggest question is whether that insurance will pay. It won’t if the tail risk is the sun blows up, but then you won’t be making claims either. The question, I think, is whether it’s sensible to buy the insurance. How do you price it? You as provider and you as buyer. If it’s sold as a claim on a stock of expected ultra-safe aspects, then it has to be expensive: you’re buying into a pool, aren’t you? That pool can’t do much because more return would put it at risk, so the expected yield should be low – unless you’re being sold fake ultra-safe stuff.
How do you as a buyer know a price? The next tail event might not affect you. It might wipe you out anyway. What if you’re hit in one area but not the area you’ve insured? Do people buy insurance for meteor strikes? How many buy trip insurance or flight coverage in case of a crash? We do all sorts of risky things everyday and yet most people don’t have much life insurance. If you have lots of money, maybe it makes more sense to buy art or jewels. Too many questions for this to be sensible.
The number of US-based institutional investors that investigated (and continue to investigate) tail-risk hedging is absolutely massive. Having very recently left a role as a derivatives specialists (listed index option-based strategies) at a $300B asset management group I can assure you that the number of accounts that we ran some form of tail-risk overlay for become substantial. Yes, the dedicated collateral to those overlays in tiny because of the efficiency of the margin requirements attached to listed index options, but the notional values they are structured to protect are substantial, and growing.
There is no such thing as a cheap put, true, and that is what many misguided clients came to us looking for. Still, the idea that the industry hasn’t evolved to the point where large asset allocators are now sophisticated enough to evaluate and implement option-based hedging approaches just isn’t correct. From targeted volatility funds to using options spreads like collars, put-spread collars, and condors to fund long put positions. These strategies have real and growing traction.
James Montier at GMO published an excellent and brief white paper titled “An Ode to the Joy of Cash” analyzing various tail risk strategies, including cash. Its a good read, with real numbers. The pdf can be found on the GMO website.
Read Full Article »
