Any resemblance between the following parable and current events is purely intentional.
Imagine an ambitious young man who wishes to become very rich, very quickly. He has sophisticated skills in manipulating data and works for a large financial firm. What is the quickest way to accumulate $6 billion dollars?
Obviously, there is no easy way to get rich. But if you are willing to risk going to jail, then you can considerably shorten the odds. So here’s my plan (kids, please don’t try this at home):
1. “Borrow” $30 billion in customer funds for a few minutes. Put $1 billion on each of numbers 1 through 30 on a roulette wheel. (If the casino doesn’t do bets this large, do the Wall Street equivalent with derivatives.)
2. There’s more than a 75 percent chance your number will hit. (A 30/38 chance, to be precise.). If you win, the casino pays you $36 billion. Quickly return $30 billion to the customer accounts and no one is the wiser.
3. Of course there is a non-trivial chance your number won’t hit, in which case your fraud will eventually be discovered and you’ll go to jail.
If you are lucky, you will be lauded as a highly talented investor: “How was this young man able to go from zero to $6 billion in such a short time? He must have found market inefficiencies. No one could beat a truly efficient market that consistently.”
But he shouldn’t push his luck. If he keeps doing these sorts of bets over and over again, then eventually he will get caught. The same media outlets that called him “the next Warren Buffett” will now claim that it was always obvious that he was a fraud.
There’s a lesson here for the Efficient Markets Hypothesis (EMH). If some people are willing to risk going to jail in order to get rich, then the number of unusually successful investors will appear to be much larger than what one would expect if markets were efficient. But this reflects cheating, not superior investment skills. Even if asset price bubbles do not exist (which is my view), reputation bubbles most certainly do exist. Some investors have reputations that are inflated far beyond their fundamental values. When those gains are achieved illegally, it is only a matter of time before the reputation is punctured.
To summarize, the EMH is truer than it seems.
READER COMMENTS
Thomas Lee Hutcheson
Jan 8 2023 at 2:38pm
Good analysis, although I doubt that SBF actually started out to commit statistical fraud, but rather that his business model stumbled into it.
vince
Jan 8 2023 at 4:45pm
EMH can have different meanings. Some say the GFC was proof EMH is BS. It depends on how EMH is defined.
Scott Sumner
Jan 8 2023 at 6:50pm
“Some say the GFC was proof EMH is BS.”
Those claims were discredited by subsequent events.
Grant Gould
Jan 8 2023 at 5:41pm
A relatively common style of fraud in the early internet was to claim to have a way of predicting the outcomes of sporting tournaments. The aspiring fraudster would create 2^n different, seemingly independent websites, each predicting one possible outcome of each of the n games of a season. Then at the end of the season they would delete the 2^n-1 incorrect ones and have a clear record of having perfectly predicted every game. On the basis of this, they could then get “investors” to buy into their next season of predictions.
One could argue that this approach using stock picks would not be very different than operating a large number of different actively managed mutual funds and then touting the winners while quietly discontinuing the losers.
In this way beyond a certain point the EMH would always appear to be violated at every specific moment, without ever actually being violated. Survivorship bias and Simpson’s Paradox will always provide the appearance of anomaly — to efficiently meet the market demand for fraud, even.
Scott Sumner
Jan 8 2023 at 6:51pm
Good example.
Henri Hein
Jan 9 2023 at 3:15pm
I suspect the Motley Fool brothers of some variant of this. They claim (or used to claim) a consistent annual 30+% return. I signed up for one of their newsletters and didn’t get anywhere close to that for those stocks. I still got good information and recommendations from the newsletter, so I wasn’t unhappy. I just found their claim dubious. I theorize they maintained multiple portfolios, so they could always point to one of them to back up the 30% number.
Jim Glass
Jan 9 2023 at 3:28pm
One could argue that this approach using stock picks would …
Oh, there was plenty of this done using stock picks and market timing calls back in the days when direct mail was the #1 form of advertising to investors. Start with a big fat mailing list, at every iteration the mail package says: “You’ve seen us be right N times in a row! For only $1,000…” There are very few new investment scam concepts, the details of their implementation just change with the times.
Spencer
Jan 9 2023 at 9:48am
EMH = “hypothesis that states that share prices reflect all information and consistent alpha generation is impossible”
…
All information is clearly not known. Joseph Granville is a good example.
Capt. J Parker
Jan 9 2023 at 1:13pm
I think there is one slight twist on Dr. Sumners analysis:
Initially, the bright ambitious young man discovered that 8 of the 30 numbers never come up on a certain roulette wheel. (Because crypto traded in Japan at a 50% price premium from US based trades)
The young man places increasingly bigger bets with money obtained under dubiously legal circumstances all the while thinking there was little risk of going to jail because there was little risk those 8 numbers coming up and little risk of the bets not paying off. He was right for a while.
Eventually the 8 numbers that never came up started coming up regularly (arbitrage opportunities for crypto closed – because markets are efficient) and the risks that seemed infinitesimal became very large.
Jose Pablo
Jan 9 2023 at 4:06pm
“Some investors have reputations that are inflated far beyond their fundamental values.”
Yes, indeed.
There is much less information about the investor’s skills in his/her investment outcomes that usually presumed.
And that’s because what a good investor does is to change the odds of getting a positive result. But neither the “zero skill” initial distribution of outcomes nor the “distribution of outcomes affected by the investor set of skills”, can be “observed” (not, at least, in the short run, let’s say 10 years at least. Maybe never).
Being this the case, it is almost unavoidable to assign too much explanatory meaning to the “actual investment outcome” just because this is the only variable that can be observed.
“To summarize, the EMH is truer than it seems.”
Other possible conclusion (not necessarily in contradiction with this one) would be that if you are a crook Wall Street is your place.
Jim Glass
Jan 9 2023 at 4:40pm
Good point about the EMH.
As to scammers, professionals are marked by having an exit strategy. E.g., someone capable of borrowing $30 billion of customer funds for a roulette spin can put 3% of it in a bank account in a foreign country with “friendly” law enforcement and a plane ticket to get there. If his spin wins, he repays that too. If not, he’ll be a well off local celebrity. (Say, Robert Vesco during his Caribbean years.)
Famous scammers in jailhouses had no exit strategy. Typically they either drank their own Kool Aid or, more frequently, messed up a legit business and tried to short-term scam their way out of the hole, making their holes deeper and forcing their scams bigger. SBF seems to have consumed his own Kool Aid by the keg. IIRC, Madoff was the champ of the hole-digging variety.
he shouldn’t push his luck … eventually he will get caught … the same media outlets that called him “the next Warren Buffett” will now claim…
This is valuable advice even to those who honestly win huge in the markets. If Steve Case had retired upon closing the AOL-Time Warner deal he’d be walking around today as one of the greatest CEOs of history, for the way he train-robbed Time Warner’s wealth for his AOL shareholders. Instead, perhaps over a nice glass of Kool Aid, he decided to stay on to manage the combined business. Now he’s the guy who presided over the greatest train wreck merger disaster of history.
David Seltzer
Jan 9 2023 at 6:54pm
Jim, a variant of the scam that was actually a trading strategy for market makers on the CBOE, Amex, and the P-Coast options exchange. We called it the “Brazilian spread” On expiration day, a trader would short out of the money options, delta .25, position limit, 1000 contracts for $25 per contract. Buy a ticket to Brazil, go to the airport and at the close, call to see where if both calls and puts expired. If yes, go home. If no, get on the plane. A real life example, one trader clearing through a CBOE clearing firm, was position limit short a straddle Oct 19, 1987. His position at the end of the day lost $50,ooo,000. He disappeared. Four traders cost First Options of Chicago, $110,000,000.
Jim Glass
Jan 10 2023 at 7:25pm
David, thanks for the example.
It’s good to know that our top financial innovators are always out there at work in the markets … innovating!
David Seltzer
Jan 9 2023 at 7:14pm
Meant $50,000,000. Apologies.
Scott Sumner
Jan 10 2023 at 6:55pm
Jim and David, Thanks for those examples.
Comments are closed.