These are not your grandparent’s markets. Markets today are data-rich, exceedingly fast and inter-connected. Importantly, the influences of data and interactions between data can be increasingly difficult to uncover. Understanding markets has never been a simple task, but the problem has become more difficult today.
Beyond simply understanding what is moving markets, we believe it is critical to adapt as markets change. In an era of heightened efficiency, strategy crowding and speed, often the life-cycle of a trading strategy—sometimes referred to as “alpha decay”—can become shortened. In other words, even if we can figure out what is influencing market moves today, there’s an increasing chance that in the near future things will change. Building a strategy for today’s markets is no guarantee of success tomorrow.
“Building a strategy for today’s markets is no guarantee of success tomorrow.”
A simple example is in the energy sector, and specifically crude oil. A little more than a decade ago we saw the price of WTI crude oil at about USD 150 per barrel and the Organization of the Petroleum Exporting Countries (“OPEC”) held the key to global oil production. At that time, it would have made quite good sense to produce a hypothesis-based trading model that keyed off fundamental data (in the form of OPEC supply numbers) in order to trade crude oil.
Enter “fracking,” a technological revolution centered in the United States. Fracking is focused on the production of oil using horizontal drilling and a process of injecting high-pressure water or other liquid into buried rock, with the goal of forcing fissures open and extracting oil or gas. Fracking had the effect of significantly increasing the supply of oil, and importantly also had the effect of shifting the balance of supply between OPEC and non-OPEC production.
With increased global supply from non-OPEC sources, OPEC production numbers would be less influential in determining the price moves in crude oil, as the link between these specific numbers and moves in the market was significantly weakened. As a result, the model we cited earlier—a hypothesis-based strategy built to trigger off of certain OPEC supply numbers—would have started to underperform, and the alpha that the model was generating would have decayed, potentially significantly.
At this point, faced with an underperforming hypothesis-based model, the research team would have had to address the issue, and turn away from whatever other projects they may have been working on. Should risk to the model be cut in half while they try to understand why the model is not working as expected? Is this a cyclical or temporary phenomenon, or is this something more permanent? Should the model be removed entirely from the portfolio? Should a new hypothesis be formulated?
One approach to dealing with the challenge of evolving markets is to build strategies that themselves can evolve. To do this in a scalable, systematic way, one must set aside the pursuit of simple hypothesis-based models in favor of what we refer to as “learning” strategies—models that can adjust autonomously simply through the continuous evaluation of data.
“One approach to dealing with the challenge of evolving markets is to build strategies that themselves can evolve. ”
Learning strategies start with the idea that we will use history, as opposed to a particular hypothesis, to drive our positioning in markets. Here history is defined as generally years or decades worth of time-series data across a range of different data inputs, or “features,” each of which may have some influence on price movements.
A data-driven approach takes as its premise that the best approach to trying to forecast movements in markets is to look at history and search for similarities with today’s environment. In other words, if today’s environment shares characteristics with some environments we have seen over the last 10, 20, 30, even 40 or more years, we can look at what happened to market prices then (these are known results) and use that as a guide for how we should be positioned now. And while the behavior today is not likely to perfectly match history, it may rhyme well enough to provide a statistical edge.
While the strategies are built from historical data, it is not a case of “one-and-done.” The building process is iterative, and as markets evolve, the most recent data is incorporated and an expanded training set is formed. The models will process this data and evolve their structure, importantly without a human being directing these adjustments.
“The building process is iterative, and as markets evolve, the most recent data is incorporated.”
In the case of our fracking example, through the incorporation of the most recent data, the models would have likely started to sense that the OPEC inventory data was becoming less predictive, and would have likely down-weighted the importance (influence, or weight) of that data in the model…perhaps increasing the importance of other data instead.
For investors, even after reviewing excellent historical performance, the question always remains, “will the strategy work in the future, under potentially different market conditions?” We believe that the use of learning strategies holds the key to building a nimble, responsive portfolio, for today and for the future. ●
***
IMPORTANT DISCLOSURE
The information provided is accurate as of the date indicated and may be superseded by subsequent market events or for other reasons. Charts and graphs provided herein are for illustrative purposes only. The information in this presentation has been developed internally and/or obtained from sources believed to be reliable; however, neither Millburn nor the author (if not Millburn) guarantees the accuracy, adequacy or completeness of such information. Nothing contained herein constitutes investment, legal, tax or other advice nor is it to be relied on in making an investment or other decision.
This presentation should not be viewed as a current or past recommendation or a solicitation of an offer to buy or sell any securities or to adopt any investment strategy.
The information in this presentation may contain projections or other forward-looking statements regarding future events, targets, forecasts or expectations regarding the strategies described herein, and is only current as of the date indicated. There is no assurance that such events or targets will be achieved, and may be significantly different from that shown here. The information in this presentation, including statements concerning financial market trends, is based on current market conditions, which will fluctuate and may be superseded by subsequent market events or for other reasons. Performance of all cited indices is calculated on a total return basis with dividends reinvested. Neither Millburn (nor any author other than Millburn) assumes any duty to, nor undertakes to update forward looking statements.
The performance and other information is based on that which is available as of the date of this report. Any markets, models, leverage, portfolio weights and other data or statistics described change over time, but are accurate as of the date indicated herein. This information is not an offer to sell any product or a solicitation of an offer to invest or open an account (an “Account”) and must be supplemented by a disclosure document when considering an investment. An Account may be opened only after receipt and review of a disclosure document and execution of certain agreements. An Account disclosure document contains important information concerning risk factors, conflicts of interest and other material aspects of an investment; this must be read carefully before any decision whether to invest is made.
Commodity interest accounts are illiquid, speculative, employ significant leverage, and involve a high degree of risk. Commodity interest products involve high fees. Please see the disclosure document for a detailed description of these and other "Risk Factors" and "Conflicts of Interest." There can be no assurance that an Account will achieve its objectives.