This article combines thoughts from an article by Howard Lindzon and my analysis on the performance of random trading. I argue that monkeys win in this market because the bulk of quant trading is worse than random.
Howard wrote today in his article “The Monkeys Win…Especially The Ones Who Understand Growth“:
“I think the funds have gotten too complicated and fancy with formulas and have missed the biggest shift in the markets…how growth is valued. ”
There is strong evidence that the infatuation with formulas and machine learning especially is hurting hedge fund performance. I have already provided evidence that more than half of machine learning predictions in the numer.ai website are worse than random.
Yesterday in a tweet I provided evidence via simulations that more than 70% of random long-only SPY traders are profiting year-to-date and 20% of those random traders who just toss a fair coin to enter and exit the market are making more than buy and hold. How many funds are making more than the buy and hold year-to-date?
— Michael Harris (@mikeharrisNY) September 22, 2016
Howard is correct: All the fancy math and formulas have not helped hedge funds. Actually, most of their strategies are worse than random and as a result, random traders look good. This should be a wake-up call for hedge funds managers. Knowing R and Python is neither sufficient, nor necessary for making profits in the markets.
But we have reached the pathetic level of ignorance at which someone who knows Python or R immediately qualifies to develop strategies for managing hard-earned savings of people.
Careful analysis will show that the bulk of these strategies suffer from data-mining bias and some from look-ahead bias. It is disconcerting that strategies with look-ahead bias have even passed peer-review for publication in journals.
And random traders are laughing…laughing hard…because tossing a coin has proven a much robust strategy than those naive algos the kids out of school are developing. It’s not their problem; it’s the fault of their management.
If your odds of winning by tossing a coin are 2.5 to 1, your chances of beating the majority of hedge funds are close to 100%.
When you realize that the strategy that invests your hard-earned money was developed by a kid out of college with excellent knowledge of computer programming but without the slightest idea of how markets work, then your only choice is to run, run away and save whatever you can.
Even worse, when you realize that a physicist (remember LTCM?) or even math professor with no concept of the market and no trading experience, regardless of age, who thinks in terms of Newton’s Law or Quantum mechanics, is developing the strategies to manage your money, you should run as fast as you can.
In the markets, notions from logic, sciences, probability theory and all other sciences are challenged daily, refuted, and then confirmed, only to be refuted the next day.
My opinion is that investors should look for an honest and experience traditional fund manager who understands asset diversification and allocation.
Traders should avoid things they do not understand and buzzwords such as machine learning. Machine learning has many good and successful applications outside of trading. When it comes to the markets, effectiveness is challenged by non-stationarity and regime switches. There are few who are successful with it but it is because they understand markets above all. And this is the key to success…
Technical and quantitative analysis of Dow-30 stocks and 30 popular ETFs is included in our Weekly Premium Report. Market signals for longer-term traders are offered by our premium Market Signals service.