Quantitative Trading

I trade with my private capital only and am not currently interested in managing anyone’s money.  But, I find the process fascinating and want to share a bit of the journey with anyone interested.

Financial markets are wonderful.  They can simultaneously create and destroy wealth.  They are not fair.  They favor those who already have capital.    If you choose to participate beyond buy/hold/re-balance, you will be up against some of the smartest and most ambitious people in the world.  Exciting!  If you haven’t read Flashboys by Michael Lewis, you should.  It’ll help you realize your place in the pecking order and just how far behind you are before you start.  Then, make sure to read Fooled by Randomness by Nassim Taleb and Nate Silver’s The Signal and the Noise.  It’ll help you fight the urge to see patterns where they don’t exist.  Lastly, brush up on your math.  Machine learning techniques are based on a mathematical foundation and back-testing only makes sense in the context of statistical significance.  That said, the efficient-market hypothesis is only mostly true.  There are opportunities every day to exploit deviations from efficiency.  Here’s a high level architecture of how I do it:

quantArch

A few key takeaways:

  • I don’t hold positions overnight
  • I strive for consistent small positive daily returns (base hits, not home runs)
  • This was a lot harder than I imagined it would be
  • It requires constant adult supervision and model tuning
  • I trade 10-20 different equities every day
  • Machine Learning cannot find patterns that don’t exist
  • The signal is weak in time series of transient data – particularly financial market data
  • Survivorship Bias shows up in some subtle ways
  • Technical analysis doesn’t work
  • Fundamental analysis eventually works – sometimes
  • Markets can stay irrational longer than you can stay solvent
  • Sometimes you lose
  • Manage your risk
  • Never give up