One of my mentors in artificial intelligence( AI) always says that with modern machine learning technologies you can find almost any answer but the hard thing is to ask the right questions. That principle certainly applies to crypto-assets. As a new financial asset class, crypto-tokens are, more often than not, evaluated using traditional metrics based on price and volume but we can do so much more. In a data-rich universe where blockchains and exchange data generates billions of data points, we can certainly find all sorts of fascinating patterns and factors that explain behaviors in crypto-assets. The hard thing is to know what to look for.

Who would you describe Ethereum? Based on the price? As a blockchain? Based on the decentralized applications(DApps) built on top of it? Understanding the behavior of crypto-assets starts by asking the right questions. A lot of our questions about crypto-assets are targeted to determine relationships with price movements (what else ?? ) but the analysis could go so far beyond that. Very often, non-obvious correlations provide the greatest insights.

During World War II, the Pentagon assembled a team of the country’s most renown mathematicians in order to develop statistical models that could assist the allied troops during the war. The talent was astonishing. Frederick Mosteller, who would later found Harvard’s statistics department, was there. So was Leonard Jimmie Savage, the pioneer of decision theory and great advocate of the field that came to be called Bayesian statistics. Norbert Wiener, the MIT mathematician and the creator of cybernetics and Milton Friedman, future Nobel prize winner in economics were also part of the group.

One of the first assignments of the group consisted of estimating the level of extra protection that should be added to US planes in order to survive the battles with the German air force. Like good statisticians, the team collected the damage caused to planes returning from encounters with the Nazis.

For each plane, the mathematicians computed the number o bullet holes across different parts of the plane (doors, wings, motor, etc). The group then proceeded to make recommendations about which areas of the planes should have additional protection. Not surprisingly, the vast majority of the recommendations focused on the areas with that had more bullet holes assuming that those were the areas targeted by the German planes. There was one exception in the group, a young statistician named Abraham Waldwho recommended to focus the extra protection in the areas that hadn’t shown any damage in the inventoried planes. Why? very simply, the young mathematician argued that the input data set( planes) only included planes that have survived the battles with the Germans. Although severe, the damage suffered by those planes was not catastrophic enough that they couldn’t return to base. therefore, he concluded that the planes that didn’t return were likely to have suffered impacts in other areas. Very clever huh?

What Wald’s story teaches us is that, no matter how sophisticated the mechanism for analyzing data, asking the wrong questions will get us nowhere. When comes to an unknown universe such as crypto-assets, that premise holds truer than ever.

We know the price and “semi-fake” volume of any crypto-assets but what other factors do we need in order to understand the behavior of this new asset class? If we use price as the driving factor for most of our questions, then it is key to understand the types of relationships we can extrapolate between prices and other factors. At a high level, there are some characteristics about the relationships between two variables that are important.

· Correlation vs. Causality: When we see a factor that fluctuates similarly to price, we tend to assume that it can be use as a predictor or price movements. While many factors might have obvious correlations, that seldom means that there is a causal relationship between the two. Let’s take a time in the market in which the prices of gold and Bitcoin are both trending upwards. While the correlation might be useful to make headlines on CNBC or Bloomberg, it might be far from explaining a causal relationship between the two. A very simple explanation could be that higher levels of volatility in US equities is causing investors to move some of their positions to Gold while a deacceleration in the Chinese economy is increasing investments in Bitcoin. If the causal factors change, then the “apparent” correlation will disappear.

· Linear vs. Non-Linear: When we think about relationships between price and other factors we visualize them as linear correlations that “co-move” together. However, some of the most fascinating patterns in financial asset investments are based polynomial, exponential and many other non-linear representations. For instance, a movement in the price of Bitcoin could be attribute to the expiration of many future contracts a few months down the road and an increase optimism in the guidance of tech companies during earnings season.

· Uni-Factor vs. Multi-Factor: We are constantly tempted to find a one-to-one correlation between a given factor and price. However, many price movements can be explained by complex linear and non-linear combinations of different factors that are far from obvious to the human eye. For instance, a price fluctuation in Bitcoin can be a combination of an increase in new investors joining the network and the fact that the price is moving in areas closer to the position of a large number of investors.

Now that we understand the types of relationships we can extrapolate between price and other factors, we can start thinking about the type of relevant questions we can ask in order to understand the behavior of a crypto-asset. In our analysis, we believe that most of the relevant questions about crypto-assets can be grouped in five main categories:

Financial questions attempt to describe the financial behavior of an asset. Some non-trivial questions in this area:

· Are investors making money or loosing money?

· How up or down are investors in their respective positions?

· Is the number of large transactions increasing or decreasing?

· Is trading happen within new investor or historical investors?

· …..

Crypto-assets operate in networks which dynamics are incredibly relevant to price movements. Relationships in crypto-networks describe growth patterns as well as unexpected movements of crypto-assets. Some relevant questions in this area:

· How many addresses are capitulating?

· Are new addresses joining the network?

· ….

Understanding counterparties has been an elusive goal of all financial asset classes but blockchains bring a unique dimension to this problem. For the first time in the history of finance, the behavior of individual investors is recorded in public ledgers. Some relevant questions related to ownership could be:

· Is a crypto-asset over-concentrated?

· Are large investors buying or selling?

· Are minority investors buying or selling?

· Are investors long-term or active-traders?

· ….

Complementing the previous point, crypto-assets offer a wide canvas to understand the psychology of investors. While we can’t predict patterns for individual investors, there is enough information to extrapolate relevant trends at the token level. Some interesting question in this area:

· Are token holders overly confident?

· Do they follow trends?

· Are they averse to loss and unlikely to liquidate their positions?

· …..

Regardless of the non-obvious correlations, crypto-assets are part of the broader financial markets and is influenced by marco-economic factors as other asset classes. Understanding the relationship between crypto-asset and macro-economic factors can generate questions such as the following:

· Is money moving from Chinese equities into Bitcoin?

· Are there visible correlations between micro-cap stocks such as the Rusell 2000 index and crypto-assets.

· ….

Understanding and predicting the behavior of crypto-assets is a fascinating and data-intensive exercise. As you start digging deeper into the behavior of crypto-assets you will confront a puzzling fact: asking the right questions is more important than getting the right answers.