The Research HUB: Regime Alignment

Momentum

Misinterpretation and being misunderstood are inherent in any domain. It is one thing to withhold judgment and it is another to misinterpret something. Part of the reason we build quantitative or mental models for any domain of life is to decrease cognitive load and decrease the margin of error for interpreting reality.

For example, you don’t typically do long and complex math problems without a calculator. The calculator allows you to move to higher levels of conclusion. This same dynamic exists in financial markets or any business. You build out processes for quantifying your environment and running operations.

Markets:

When we think about markets, it is important to quantify regimes correctly so that you don’t let moments of misinterpretation misguide you. Two articles I have already written on this topic can be found here:

• The Research Hub: Risk On / Risk Off Regimes

• The Research HUB: Intraday Trading Primer

There are specific ways to correctly interpret each section of the economy and financial markets but price action is one of the most important. While there is a significant amounts of academic literature around these ideas, I am going to explain them simply and illustrate how they directly connect to my trading strategies.

Fundamentally, price action can exhibit momentum or mean reversion in either its left or right tails. When I think about a market view, I don’t fundamentally think about bullish or bearish. I think about the left or right tails and how either momentum or mean reversion can occur across these distributions.

Momentum:

What is momentum? Simply put, it is the persistence of a time series in a single direction. Two great intro papers on this can be found here:

• https://www.aqr.com/Insights/Research/Working-Paper/Factor-Momentum-Everywhere

• https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3489539

These papers show ways you can quantify momentum but let’s use something simple like: If price > 50D moving average = positive momentum regime.

Simply put, momentum is the persistence of price action in a single direction. “Momentum” says that momentum can predict further momentum in the same direction. This is not a hard and fast rule but it is a principle with statistical significance and can be used as part of the process for making a decision.

Mean Reversion:

Mean reversion functionally asserts the opposite of momentum. It says that price will continually revert back to the mean. For example, you might use Bollinger bands to illustrate the price continually reverting back to the mean after a 2-3stdv move.

Let’s dig into something tangible. Here is the NVDA chart with the Bollinger band and moving average (in red). If my strategy is betting on momentum, then I could use the moving average to moderate my actions. Inversely, if I want to bet on mean reversion then I could use the Bollinger band. Let’s say I shorted NVDA every time it hit the top end of the bollinger band range. How would that work? Not great because I am betting on mean reversion when momentum is happening.

However, the opposite can be true if look at a different period of time in the NVDA chart. During this time, the market was rewarding mean reversion

You can begin to see that the market rewards different outcomes at different times. For all of my trades, they fall into one of four categories:

• Betting with momentum

• Betting against momentum (you must have an informational edge for this!)

• Betting with mean reversion

• Betting against mean reversion

These categories do not have anything to do with being bullish or bearish, they have to do with the TYPE of price action that is likely to occur. This is critical for execution, risk management, and falsifying a thesis.

Tangible Examples:

Now if you are feeling a little overwhelmed, that’s good. Growth happens in difficulty and challenge. My goal is to provide you clarity so that success is on you and not a poor explanation on my part.

Let me start with, each of the categories I noted above are tools for managing uncertainty. There is no magic solution to all things. I view these ideas as directly connected to how complex adaptive systems fundamentally function (read this book on it: Link).

Let me run through a couple of real-life examples and then we will go through market examples:

• Betting with momentum tangible example: If you are running a company that produces goods or services for the economy, how do you make decisions about investment, inventory and labor? For example, if sales have been trending up, should you double down on investment, inventory, and labor or should you begin scaling back? Do you see how this idea of momentum is directly connected to any decision under uncertainty? The question is, how much of the momentum in x trend should I align my decisions with? (you can also use this framework to analyze companies’ decisions).

• Betting against momentum tangible example: If you are running a company and the entire sector is trending down in its sales and output, what do you do? For example, in 2008, do you start buying houses or start a homebuilding company? All the momentum is pointing to the downside so do you bet against this? There is always a reason for momentum moving in a specific direction so you need to establish that the current driver of momentum will stop as opposed to having a view about what “should” be happening.

• Betting with mean reversion tangible example: let’s say you are running a commodity business that has a higher degree of seasonality in it due to the weather. There will always be fluctuations to the upside or downside and the question is, do you bet that these forces revert back to their mean?

• Betting against mean reversion tangible example: If you are running a commodity business that typically has a lot of oscillations that revert to the mean, is there a scenario in which you make decisions that bet on things to shift from mean reversion to momentum? Functionally, the absence of mean reversion.

Now these principles are likely incredibly intuitive for you already but quantifying them is HOW you are able to reduce cognitive load and move to higher-level decisions.

Let’s go through some market examples:

Let’s go back to our moving average example. Let’s say you are betting with momentum where you buy when price > 50D MA. This would work great during momentum regimes. However, this strategy would get chopped up and lose money during a mean reversion regime (see yellow area).

Here is another example. Let’s say in the mean reversion regime you get long or short at the BB extremes. This would produce exceptional returns and even more trading opportunities than the single momentum trade. However, the momentum trade can have greater persistence.

Here is where it gets complicated:

Now that we have a basic understanding of momentum and mean reversion as it occurs on the right or left tails, let’s get down to why this isn’t as easy as it looks.

Problems to Solve:

• Momentum and mean reversion are typically always occurring across multiple timeframes. For example, price could be exhibiting momentum on an intraday basis but mean reversion on a weekly basis. The same could be true on a weekly and yearly basis. How do you solve for the optimal timeframe especially if the market regime influences this?

• How do you quantify momentum and mean reversion? While the moving average and Bollinger bands I used above are helpful in illustrating the principles, they are incredibly simplistic and need to be refined considerably to provide a technical edge. In my view, this is how you quantify price action to begin developing a technical edge as opposed to drawing technical trend lines.

• How do momentum and mean reversion outcomes connect to the underlying dynamics in macro or the fundamentals of a company? See my notes here on this:

Once you begin to connect your regime analysis with momentum and mean reversion outcomes then you are able to optimize trade selection exponentially. However, accomplishing this is incredibly difficult and literally why firms like

Mythic Market Research

exists to do consulting in these types of scenarios on the price action side.

The final thing I will say is that if you are trying to increase diversification in a portfolio, implementing strategies across BOTH momentum and mean reversion significantly reduces correlation risk. However, implementing this across all major assets is incredibly time-intensive (See the

SpearPoint Equity Alpha

for tangible examples).

But as you know, nothing exceptional takes a small amount of time.

If you want to dig into these ideas more, check out all the educational articles I have written here or just send me an email:

In the information age, you simply need to be at the right place, at the right time, with the right information to succeed