Transtrend started as a research project in 1989. We had bought data and computers, hoping something could be done with that. So, that’s what we tried. We dove into the data. As a result, Transtrend’s Diversified Trend Program came into being.  

Our starting point was a treasure trove: up to 20 years of daily market data — open, high, low, close, volume, and open interest — across 35 different futures markets. Storing all that data required a cabinet full of large, removable disk packs. The total? Less than 250 MB. Today, Transtrend stores more than a million times as much data, and our computing power has also grown exponentially.

Are our investment decisions now exponentially more informed as well? Sadly, not. The main reason is that it’s not facts, as recorded by data points, that drive markets. It’s stories.

Storytellers

We didn’t always believe this. Back then, the investment world was split into two camps: technical versus fundamental analysis. At top investment banks, using technical analysis was a firing offense. Markets were driven by supply and demand, not by trendlines or head-and-shoulder patterns. Technical analysis was dismissed as astrology for finance. On the other side, technical analysts dismissed fundamental storytelling as the Delphic oracle. Prices formed the only objective reality. All information was compressed in that. We largely sympathized with that view.

To be honest, this distinction between data and stories was never as absolute as often proclaimed. Our ‘only looking at prices’ is itself a story. For example, no futures trader wants to take delivery of tons of soybeans. So, even the most fundamentals-averse futures trader takes delivery dates into account. One could argue these are just other objective data points, but the choice to avoid delivery is a fundamental story. And data is needed to uphold that story.

Another oft-repeated CTA story is: “We trade the same size in all markets.” These managers aim to avoid overcomplicating their approach with various subjective assumptions. But what exactly does ‘same size’ mean? Same number of contracts? Same notional value? Same margin? Same risk? It could truly be anything! This simple example illustrates that data alone is meaningless. It must be translated in some way to become useful — as part of a story. Also when that story itself sounds extremely simple like ‘same size’. However, this story resonates, even when it’s unclear what it means in terms of objectively measurable data. Or because of that?

Stories can be powerful — even more powerful than data. This was something we had to learn over the years. In our early years, we naively believed allocators would only look at the numbers: a sufficiently long track record, strong returns relative to risk — measured by Sharpe ratio or some other metric — and low correlation to other investments. But that’s not how money really flows. Money largely follows stories.

One of today’s success stories is AI. Whether or not AI will ultimately deliver all its supposed benefits is irrelevant to its current market impact. Companies mention AI in their earnings calls, and their stock rallies — even if the actual numbers disappoint. The story is what matters.

As quants, we may be skeptical of this phenomenon; we may dismiss these stories as fairy tales. But the truth is, it’s very hard to make money by betting against the ultimate forces that drive markets. And these forces are stories — not facts, not data. Stories can outlast the means of even the most accurate quant.

The academic foundation

Our first taste of ‘more data ≠ more informed’ came in the early 2000s, when we started collecting tick-by-tick data. For us, high-frequency data meant lots of data without a long history; precisely what we needed for our correlation analysis. The correlation structure between different markets isn’t linear. Modelling that reliably required a lot of data. However, by using daily data this implied ‘a long history’, which conflicted with our conviction that correlations change over time. Eureka!

By choosing for high-frequency data to solve this problem we implicitly assumed that market dynamics are the same across timeframes. This is itself a story — and a very popular one among technical analysts in those years. It formed the foundation for viewing price patterns as fractals.  Unfortunately, we soon came to the realization that this assumption was incorrect. Market dynamics — including their correlation structure — are not constant, but instead vary across different timeframes. Behavioral finance helped us understand why. So, we stopped searching for a solution in that direction.

This exemplifies the academic approach: form a hypothesis, test it on real-life data, and reject it if the data doesn’t fit. Which includes: reject the story. If you’d asked me back then about a potential explosion of available data, I would have replied that it would make it easier to bust false stories. Turns out, I was wrong.

As I was also wrong to believe that flat-earthers would extinct once information delivered through satellites circling our planet would become our daily feed of perception. I overlooked the fundamental principle of inflation: an exponentially increasing amount of data diminishes the value of each data point. More data makes it easier — or even necessary — for people to ignore most of it. It encourages us to do our own research on only the data that supports our chosen story.

The flat-earthers are not an exception; we all do it. We all embrace stories, including us. Our skilled analytical approach doesn’t make us superior. As quants, we are just as susceptible to stories as anyone else — perhaps even more so, because we love abstraction and tend to be less distracted by ‘messy’ details.

This tendency of ours isn’t new. Around 500 BC, the Pythagoreans believed that all phenomena in the universe could be reduced to whole numbers and their ratios — what we now call rational numbers. Irrational numbers couldn’t exist. But then one of them, Hippasus, irrefutably proved that √2 wasn’t rational. Did they abandon their story? The Pythagoreans weren’t ready for that. Instead, they kept the discovery secret. And Hippasus was killed.

Stories without facts

The Pythagoreans preventing Hippasus from making his discovery public was comparable to the Roman Catholic Church forcing Galileo to recant his heliocentric theory. And hereby we’re entering religion. To understand what drives markets, we must understand what drives people. And we cannot seriously describe what drives people without addressing religion. Let’s not treat religion as the irrational numbers were once treated!

There are certainly Christians whose faith is rooted in accepting specific biblical accounts as literal facts — such as the sun circling the earth and Noah’s ark saving one pair of every species. Just like there are people who reject Christianity by ‘academically’ rejecting one or more of these facts. But for most Christians, these facts are not taken literally. Their faith is grounded in principles such as loving God and loving their neighbor as themselves.

Other religions are not essentially different in this respect. Likewise, most people who aren’t religious still embrace the story of love. Many people of my generation will have seen Pretty Woman. Was that movie really about a corporate raider who tried to acquire a shipbuilding company? What if it had been a railroad company? Or are these irrelevant details?

I would say that the film was compelling because it touched on a universal truth. This very same truth — with completely different ‘facts’ — has been portrayed in many other equally moving films. In cartoons, the same story has been told using animals, trees, or even unidentifiable creatures. No, Shrek isn’t real, but the story is.

A true story isn’t a collection of facts. It exists independently of facts — sometimes even despite them. Yet, to tell a story or to present it in a movie, we can only list or display a series of facts. In this sense, facts are just the pixels that paint the story. What these pixels represent is essentially irrelevant; the same pixels can paint a completely different story, and the same story can be painted with different pixels. 

Facts exalted above stories

Often, some of these essentially meaningless pixels become sanctified. They turn into symbols. In religion, they manifest as rituals: attending church twice every Sunday, wearing traditional clothing. In law, it’s the elevation of the letter above the spirit. Such symbols share a key trait: they are objectively measurable. And that’s also why we quants are so susceptible to this pattern.

It happens everywhere. Take one of the most famous Dutch books: “The diary of Anne Frank”. Most people will know what it’s about; a moving story that should never be forgotten. That’s also why the ‘Achterhuis’, the place where Anne and her family hid, has been preserved as a memorial and museum.

One could argue that the Achterhuis became a symbol of (the horror of) genocide. A few decades ago, it turned out that this status extended to the chestnut tree that Anne mentioned in her diary. In 2007, when the tree was in poor condition, and experts recommended removal, protests erupted as if Anne Frank herself would be deported again. I thought I knew what Anne Frank’s diary was about. Suddenly, it was about a chestnut tree. Until a 2010 storm ruthlessly ended that story. Or did it merely fell a symbol?

But stories will prevail

This is the world we live in. It’s not just a collection of facts for us quants to measure and analyze. It’s a world driven by people — people who adopt and embrace stories. Facts are necessary as pixels that paint those stories. Some of these pixels are elevated to symbols, worshiped by the followers of the stories. And to reach that status, these facts themselves don’t even have to be true. 

Successful politicians understand this dynamic. Every now and then, a technocrat emerges, leading a political movement with only numbers. Fresh out of university, I believed this was the only proper way to lead a country. Now I know better. Technocrats don’t lead people — they annoy them. People are led by stories. And the most successful politicians jump on the bandwagon of stories that are already unfolding. We can only hope these stories are mostly factual. But for the course of future events, that makes no difference.

The storm that felled Anne Frank’s chestnut tree wasn’t human. But our economy is.  And isn’t that economy — at least, the way we quants look at it — a forest rich in chestnut trees too? Don’t we, time and again, elevate numbers? One moment it’s credit ratings, the next it’s CPIs, or job numbers. Don’t these numbers only have meaning as long as we give them meaning? But what happens when only we quants worship such data? What if the people in our societies — the people who ultimately drive our economies and vote in elections — what if our societies lose faith in these symbols?

When President Trump claims he will cut drug prices by 800, 1200, 1500 percent, such numbers make no sense from a quantitative perspective. But persistently trying to counter such numbers with ‘better’ numbers is equally pointless. It’s like a chess player explaining the position of the king and the queen to a poker player. They may use the same words, but they talk different symbols.

We quants could dismiss the quantitative skills of politicians, but that won’t improve our understanding of markets. Maybe we should realize that our facts and numbers are not the cornerstones of our societies. But is that really new? If we look back just a little further than our own academic confirmation, we might realize that we are the offspring of a quant age that lasted barely half a century  — a footnote in the hundreds of thousands of years of human civilization united by shared stories.

If stories, not data, drive our societies and drive our markets, why do we quants base our decisions on data? Answering this question could be existential.