Adam Robinson’s Game Theory Approach to Markets

Distilled from his interview with Shane Parrish on the Knowledge Project

Markets are smart

When people or in disagreement with prices or confused they are in denial or are missing something from their model

Dunking on fundamental value investing

  • Relies on Ben Graham’s undefined notion of “intrinsic value”
  • It is defined by “the value justified by the facts”. This is a meaningless definition. Like “gravity is when things go down”.
  • Thinking fundamental investing works is hubris. You must believe:
    1. There is a true value
    2. You can ascertain it
    3. Others will come around to your view in a reasonable timeframe
  • What about Buffet and Munger?
    • They hold things forever.
    • They are geniuses.
    • It is a stretch to attribute their success to this idea of fundamental investing.

Dunking on technical analysis

  • Exercise in confirmation bias and data mining

Adam’s approach: Game theory

  • He doesn’t try to predict market prices. He follows the smart money
  • The market is a predatory ecosystem. Books like Peter Lynch “One Up on Wall Street” give retail the illusion they can win in what is a ‘gladiatorial pit’
  • Keynes who was also a great investor described investing: “How do we anticipate the anticipation of others?”
  • What pattern of behavior have you seen that correlates with a different future?
    • People placing bets are wagers on a view of the future
    • His favorite investing book is not an investing book: 1962’s Everett Roger’s “The Diffusion of Innovation”
      • A trend at its core is the spread of ideas
      • Roger’s decomposes the lifecycle of an idea. Early adopters are ridiculed, the masses begin to come around, the idea is enshrined and seen as ‘self-evident’
  • His ordering of traders and how they express their views. Traders near the top of the order will be “right” on a lagged basis. The giant caveat is that these orderings may not have applied as strongly before the 2000s because he claims the world was different (different investment flows, presence of EU, etc). But he makes the case they still held. He looks for strongly divergent views between asset classes to make probabilistic bets on the future. He prefers this because it is the expression of bets vs say using economic statistics. You don’t trade statistics, you trade assets.
    1. Metals traders sentiment is proxied by the copper/gold ratio. They are the “Forrest Gumps” of the investing world —simplistic. They are the closest to economic activity. They are very far-sighted because of mine timelines. They have never been wrong in the past 18 years on the direction of interest rates. In September 2018, during this conversation, the copper/gold ratio implied that interest rates should be at 1-year lows instead they were at 1-year highs. He thinks the metal traders will again be right, they are just early. (9 months later as I write this, interest rates have gone back to 1-year lows!)
    2. Bond traders sentiment proxied by the ratio of LQD/IEF. Basically, credit spreads
      • When they disagree with equity traders, the bond traders tend to be right and early
    3. Equity traders
    4. Oil traders sentiment reflected in XLE vs SP500 spread. The price of oil is less reliable because of sovereign intervention
    5. FX traders sentiment reflected in commodity currency crosses
    6. Economists: Always wrong as a group
    7. Central Bankers: not in touch with the real economy; rely on models only. And economists
  • 3 Ways a Trend Can Form
    1. A stock very sharply reverses a long-standing trend. The trend needs to have been in place for a long time (long is ambiguous; he says ‘months’ or ‘years’). The stock will retrace after its sharp move but if it runs out of gas then the early adopter of the new direction are starting to win converts
    2. Parabolic moves precede a change in direction and a new trend in the opposite direction (reminds me of dynamics of a squeeze)
    3. An asset in a long-established, tight range starts to break out. The less patient hands have been transferring their position to hands that have more conviction evidenced by them willing to wade into a dead name.

His ranking model jives with how I think about trading

  • The science part of trading is the constant measuring of market prices and implied parameters.
    1. Rank which markets are the most efficient
    2. Find the parameters which are in conflict with one another
    3. The parameters in the less efficient markets that conflict with more efficient markets represent an opportunity set
  • The art part is to then investigate why those parameters are priced “inefficiently”.
    • Flow-based? Who’s the sucker? Who’s better capitalized?
    • Behavioral? Confirmation, anchoring, recency biases? Others?
    • Is there an aspect of the inefficient market that is unaccounted for and therefore not normalized for in the comparison to the efficient market?