Still Demise-ing

Baruch is a bit dazed at the flurry of interest his last post elicited, having been linked to by 3 giants of the econobloggyfirmament. I wonder what the hordes of recent visitors thinks of our little project, Bento. This blog was always conceived of as bit of a rara avis, as Spinoza-blogging is a rather esoteric activity, likely only interesting to a few predisposed minds. If however only one of the hordes starts being interested in Spinoza as a result of checking us out it will be a major triumph. In any case, welcome, hordes.

I would like to respond to the always interesting Felix at (by the way, Felix, I am just “Baruch”, no last name given), not to bite the hand that feeds, but to correct some very mild misunderstanding about my position which might have cropped up. He writes that “I make an impassioned plea in favor of humans over machines”, and (astonishingly) that I misunderstand the difference between “ontology”, “doxology” and “epistemology” — I have my terminology “in a twist”.

My point is not that fundamentalism and human-driven trading is “better” than computer-driven Quant, in the sense that one is ceteris paribus more likely to make more money with one strategy or another. Indeed, in another time where the trade was less crowded I might have bet on the Quant — 300% in a year from zero market risk is pretty compelling. As Veyran Allen implies, there is too much variability and often too little rigour in most human-driven trading. Emotion is a deadly enemy, so much so that investors with the emotion-inducing parts of the brain damaged are said to be better traders. Baruch has seen close up how a traumatic experience can affect a previously successful human-driven investment process, typically in the form of a confidence-destroying trading setback or humiliation. Humans cannot deploy the self-pitilessness of a computer. Computers can also work more effectively distributing responsibilities in a team, or rather, in a grid, while the best human-based strategy can fail if the traders bicker amongst themselves.

No, I do not plead for the flesh because I believe in its innate superiority over silicon, as Felix implies. However, I do think one strategy may have less capacity than the other. In that sense the problems of the Quants have been caused by their very success.

At work I like Nvidia, and Sandisk, and have been buying recently but enough people disagree with me that I have been taking a bath on both. Unlike Quant, Fundy analysis is qualitative, and can be particular to each individual — one man’s long is another man’s short.  Moreover, the successful trader is hard to copy. How often are we told we can “Trade Like Warren Buffett” in books (personally I would like to trade like George Soros), and how many of us actually do? Quant, however, is a strategy where one computer programmed to seek Gaussian correlations is likely to make the same trades as another. The most salient difference between human and Quant surely is that Quant strategies are more easily replicable (another may be that they trade more often, with more leverage). Unrestrained, successful Quants thus create “hedge fund beta”. It makes it more likely that that whole subsector of the market leans in a particular way — not necessarily in a single direction, Quant is mostly advertised as market neutral, remember, but congregated around the same trades, which creates the potential for one might call a “factor”- or “style squeeze”. And once you add 8x leverage you have something of the potential of a bomb, which is what went off in August.

That aside, there is one aspect to a lot of Quant investing that I think could make it epistemologically, ontologically, and even dogstastically inferior to brands of fundy investing in coming years, and that is that mean reversion may become a more dangerous assumption generally. As I noted in the previous post, I am beginning to think some retrenchement to the Great Moderation in volatility may take place over the next few years. Consider 3 arguments:

  • the anecodtal (subscription maybe required — James Cramer relays the complaints of veteran traders that old relationships which ruled certain sectors no longer hold)
  • the empirical — the inexorable increase in volatility indices (via Abnormal Returns again), itself possibly a function of the decline of the Quants
  • the big-picture theoretical — as we move into late cycle investing, certain sectors stop working, the market narrows as growth slows, and investors pay up for what growth remains (mostly tech, I hope). Large variations emerge in the multiples investors are prepared to put on stocks in different sectors. Too much money chases too few stories. It is an environment where the emergence of bubbles in individual sectors and stocks becomes more likely.

In a world of greater volatility, mean reversion will become riskier. The percentage hit-rate of pair trades based on simple relative performance (take the DUMB-DICK example in the original post) could fall from say 75% to 65%, radically changing the expected returns of a portfolio based on that strategy. Instead of snapping back, the statistical disparity in the relative moves could become larger, triggering stop-losses instead of profit-taking. Again, a strategy which instead bets on individual stocks, on mini Black Swans strongly outperforming their peers, would potentially do much better in an environment like this. Admittedly, that strategy could be a factor based, automated one, but the search for the next Microsoft tends to be carried out by humans, and in any case both these strategies would be fundamental ones. I leave aside the idea of whether, as Felix writes, there really is “precious little empirical evidence that fundamentals-based investing is any more successful than any other strategy”, as I have dealt with this topic elsewhere, and gven him a good drubbing while doing so.

So hopefully that clears any misunderstanding up, I look forward to robust pushback, and if any of you would like a good, accessible introduction to the works of Spinoza, you could do a lot worse than the wonderful Betraying Spinoza, by Natasha Goldstein. A fantastic book.


8 thoughts on “Still Demise-ing”

  1. Welcome to the econo-bloggy-firmament, Baruch.

    Now you understand–if you did not already–what those clever Japanese meant when they said that the nail that sticks up gets hammered down.

    Buckle your epistemo-onto-doxological seatbelt.


  2. I do not consider myself to have been successfully “hammered down”. In any case I think my short stay the firmament has so far been quite civil, even collegial. Pass the port and all that.

    Also, between you and me, are “doxological” or “doxology” real words? Are we sure they mean what we think they mean? In fact, are we just pretending to know so we don’t look stupid? or is it just me. . .

  3. Felix, Ken, TED, apologies. Perhaps Felix’s terminology in this case is correct. I spoke to the wife, a Greek and they know this stuff. She says it is possible to have a “doxology” as one has an “ontology”. It refers to a system of belief, to faith. Thus “othodoxy” — “ortho” meaning straight, regular (orthodontistry is making “straight teeth”), and “dox” meaning “faith” — the right faith.

    Felix and classical education 1, Baruch and too much time spent in state schools ,0.

  4. רבים ממזמורי ספר תהִלִים הם דוקסולוגיוֹת – המנוני תהילה לאל. כך גם שִׁירוֹת נוספות במקרא, כגון: “ה’, בצאתך משעיר, בצעדך משדה אדום, ארץ רעשה גם שמים נטפו, גם עבים נטפו מים…” (שופטים ה’ 4).

    Spinoza would have known it. Shame on you.

  5. >>”The percentage hit-rate of pair trades based on simple relative performance (take the DUMB-DICK example in the original post) could fall from say 75% to 65%, radically changing the expected returns of a portfolio based on that strategy.”<<

    That was exactly what LTCM died of; their ‘model’ kept telling them to keep adding on… —the spread was bound to close! But, as Keynes so famously remarked, “Markets can remain irrational for far longer than you or I can remain solvent.” Neverthless, I DO believe (I have degrees in psychology AND mathematics) that we can combine the QUANTa with the QUALs!

    After all, even most physical models recognize boundary conditions wher the rules do not apply.

    P.S. If all else fails, there is always WIKOPEDIA (if you don’t like the definition, you can always change it.)

  6. MIT joke paper on “You might a quant when….”

    Interesting, pithy writing Baruch!
    I was once a Trader at the bank with0ut a Prince. I left to trade for my own account. Needless to say, I did everything a** backwards, (emotion and anxiety flourished) and traded 1 account straight in to the ground. After regrouping I took used a more systematic approach, all computer generated ideas, this worked will until I made some big unexpected mistakes. I’m guessing a few of those quants were on the winning end of my trades. Luckily, I kept a journal and have since created an extravagant set of personal rules to apply to what little capital remains in my account.

Comments are closed.