“Hundreds of studies have shown that wherever we have sufficient information to build a model, it will perform better than most people.” – Daniel Kahneman
Ray Dalio and Bill Ackman are arguably two of the greatest hedge fund managers in the business today. Last week they shared the stage at an investment conference to discuss their respective investment processes. Dalio gave Ackman some interesting insights into his quantitative process at Bridgewater Associates during their discussion (courtesy of some notes from ValueWalk):
Dalio continued with the advice that you should write down your method so it can be back-tested, he feels that everything can be analyzed and quantified.
99% of the time he agrees with his quantitative strategy, the 1% of the time when he disagrees with the machine he realizes in retrospect that the machine was right 66% of the time.
Dalio is one of the most brilliant investors alive, but he’s wrong 2 out of every 3 times he tries to override his quantitative investing model. I think this is a great lesson for those investors that plan on using their gut instinct to make investment moves. The majority of the time following a rule-based approach is superior to making ad-hoc decisions.
In Thinking, Fast and Slow Daniel Kahneman discusses how study after study have proven that simple algorithms used for making decisions tend to beat experts that try to use their intuition in a wide variety of fields – medicine, business, investing and more. Experts are overconfident in their abilities, while models are not. Models are disciplined while the experts tend to let their biases blind them to potential errors in their line of thinking.
Tobias Carlisle and Wes Gray share an interesting case study on the problems that come about from making changes to a simple model in their book Quantitative Value. They looked into Joel Greenblatt’s Magic Formula stock-picking system. The Magic Formula is a simple screen that looks for quality stocks trading at cheap valuations, something Greenblatt and others have proven to be a winning combination.
Over a two year period from 2009 to 2011, Greenblatt’s firm created what turned out to be a very telling experiment created by the options presented to their investors. They offered clients two choices: a “self-managed” account or a “professionally-managed” account. The self-managed accounts would choose their own stocks using the Magic Formula list generated by Greenblatt’s screen. In the professionally managed account the stocks would be chosen automatically for them by following the model.
Greenblatt found the self-managed accounts slightly underperformed the market (59.4% vs. 62.7% for the S&P 500) while those accounts that were automated returned 84.1% after all expenses, handily outpacing the market. Adding a discretionary component to the model hurt the performance of the self-managed accounts for a few reasons. First, these investors avoided the best performing stocks. The best performers tended to be the most depressed stocks, but investors generally know why stocks are depressed so they avoid them at the time. Second, investors tended to sell stocks after periods of poor performance and buy stocks after periods of good performance, the opposite of a value strategy. Interestingly enough, it was the discretionary account that made no changes that performed the best:
Perhaps Greenblatt’s most interesting data point comes from the best-performing self-managed account. It didn’t do anything. After the initial account was opened, the client bought stocks from the list and didn’t trade again for the entire two year period. It seems even doing nothing outperformed all the other active self-managed accounts.
For most people it’s hard to believe that an automated system could make better decisions than their own gut instincts. But having a rule-based approach doesn’t mean investors have to come up with a complex black box trading system. Automation in the investment process is all about making level-headed decisions ahead of time. If you’re able to create if/then rules to guide your actions you can systematically keep overconfidence and overreactions out of your investment decisions during difficult market environments.
Obviously, investors still have to make some decisions along the way even when implementing a model or a system. Life isn’t always formulaic as circumstances change and updates have to be made to your investment plan. The point is that a rules-based, systematic approach relieves you from having to make decisions under stressful situations, something the majority of investors are terrible at.
Taking emotions out of the decision-making process will almost always be the right move because most of the time it’s difficult to see our own behavioral biases.
The Curious Case of John Hussman