We continue our blog series: Market Musings, Volume 1, Edition 16, giving our (hopefully not too random) thoughts on recent goings-on in the markets. Today, we present Don't Override the Model and A Few Thoughts on the Crypto Stock Mania.
1. Don't Override the Model - An interesting article on quant hedge fund Renaissance Technologies (RenTech) founder James Simons appears in the most recent issue of the New Yorker (link here). Most of the article centers on his philanthropic efforts to advance science, however the following tidbit was included which sheds light on his secret investing sauce: "Beyond saying that [an algorithm he had recently come up with] didn’t work, he wouldn’t discuss the details—Renaissance’s methods are proprietary and secret—but he did share with me the key to his investing success: he 'never overrode the model.' Once he settled on what should happen, he held tight until it did".
How good have RenTech's returns been? During the 20-year period from 1994-2014, according to Wikipedia, RenTech averaged a 71.8% annual return in its flagship Medallion Fund. This equates to about a 50,000X overall return during the period. $10,000 invested in the fund at inception would have turned in $502,000,000 gross. Not too bad. (An equal amount in the S&P 500 in mid-1994 would have only become $59,300 if left in to the present, if one ignores dividends.) Clearly it pays to have brilliant people, such as math genius Simons, handling one's money.
The idea that one should never "overrule the model" could also apply to non-quant funds (in our opinion). All too often portfolio managers second-guess themselves and prematurely reverse course on an investment (in effect, overriding the models that were used to justify the investment in the first place), usually resulting in mistakes and huge opportunity costs. For example, imagine a long position originally established at $20/share, which the portfolio manager believes is worth $35/share at the time the position is opened. Next assume that the stock trades down to $12/share on supposedly "bad news" or general negative sentiment (e.g., everybody on CNBC and social media is bad-mouthing the stock). Psychologically, this is tough to take, especially if the portfolio manager's clients are looking on and wondering what the heck is going on and why their money is invested in such a hated company. The manager very likely will be pressured into thinking he or she made a huge mistake and that it might be good to trim the position to "reduce risk". Yet often in these situations the very instinct which provides emotional relief if acceded to ("cutting losses and thereby reducing risk") is the exact opposite of the correct decision (and in reality dramatically increases the key risk, namely the risk of permanent capital loss); and more often than not, increasing exposure to the position in question actually reduces risk. So instead on trimming at $12/share, the manager might actually be much better off doubling down at $12/share, thereby lowering the position's cost basis from $20/share to $15/share. Obviously, the lower the cost basis the less risky an investment is, all other things being equal, since the odds of incurring a permanent loss of capital are thereby lower. To put it another way: If the model used to originally underwrite the position showed that buying at $20 was correct, wouldn't it be even more correct to buy at $15 (or $10, or $5)?
To illustrate a real-life example, in 2015 and 2016 we established a long position in Michael Kors (ticker KORS). Although our small original purchase price was $60/share, we bought as low as $38/share and were able to establish an overall cost basis of $46/share. As a partial hedge against potential deterioration in the retail sector generally, we also shorted Coach, now known as Tapstry Inc (ticker TPR) at $36/share (this position was only about 1/3rd the size of the KORS long). Our reasoning was that both KORS and TPR were very similar companies, yet TPR traded at a high teens multiple while KORS was inexplicably assigned a ~10X multiple by the all-knowing market. For a long period (at least, it seemed long when things were going against us) this pair trade did not work. KORS shares floundered while TPR stock went up and up. Eventually we lost faith and closed out both ends of the trade at a significant loss in early June 2017 (when KORS was around $34.50/share and TPR around $46/share). In effect, we overrode our model due to sustained emotional trauma suffered in seeing both positions go against us for so long. Importantly, nothing fundamental had occurred that could reasonably have led us to believe that the original thesis was incorrect.
Obviously, in hindsight our decision to close out at a loss appears to have been a huge unforced error. Below is (A) a chart showing both stocks from inception of the investment to the present (indicating that the pair trade would have worked out just fine, with KORS outperforming TPR [and remember that the TPR short was only 1/3rd of the size as the KORS long, so the downside from the TPR short would not have been too damaging]), and (B) a chart showing both stocks from the date we closed out of the investment to the present (showing that we missed a 75% up move in KORS while TPR has actually declined during the same period--not good!!!):
Next time we'll follow Mr. Simons' example and not override the model.
2. Crypto Stock Insanity - In case anyone is unaware, the recent crypto currency craze applies not just to the currencies themselves, but also to many publicly traded companies that include "crypto" in their business model. Some of these stocks have exploded several hundred percent in a day over the past few trading days. Below are a few recent examples (tickers LFIN, MARA and CRCW):
Note that the principal shareholder of the final company (CRCW), a gentleman nobody had ever heard of as of a month ago named Michael Poutre, was actually "worth" around $4 billion as of yesterday when the company's stock hit the upper $500s range (source):
Seeing this kind of nonsense almost makes one lose faith in the public markets, which in cases such as these has become a completely unregulated casino populated by hucksters, speculators, the credulous and the gullible. Whoever is buying the above three stocks at such elevated prices is virtually guaranteed to lose >95% of their investment if they hold for an extended period. Yet that is the key point--nobody buying these stocks has any intention to hold for long, rather they all hope to dump their shares at a profit on people even greedier and stupider than they are (greater fool theory). But somebody will end up stuck with all of the losses (and, likewise, somebody else will walk away with the gains). At what point, though, do we ask the following: Is society really better off with this kind of idiocy going on?
Fortunately, there is a simple way to put an end to speculative orgies of this sort. Just tax day-trading gains at a much higher tax rate than normal short-term capital gains. For example, the federal tax code could be amended to apply a 70% or 80% capital gains tax to profits from trades where the investor holds the stock for 5 or few trading days (or 10 or fewer). If this were the case, stocks such as the above "crypto stocks" would never levitate so violently, since most upside from day-trading these companies would go to the government rather than the trader (meaning that traders and speculators would simply avoid getting involved at all). With the tax code now being amended, any opportunity in this vein has been lost for the time being. Hope springs eternal, though, that one day such an "anti-manipulation / speculation / greater fool game theory" prevention measure will be enacted.