The use of alternative data for cryptocurrency trading is not new — in fact it’s sometimes the only data available, but the trend to rely on the “alternative” has spilled over into the financial markets as hedge funds search between the couch cushions for any edge they can find.

Theodore “Ted” Fillmore was a precocious child and would cause his parents no shortage of headaches whenever they brought their rambunctious 5-year-old to a friend’s home or to a party.

By the end of any social gathering, Ted’s parents would struggle to find where he had vanished to, by which time he’d already found his way through the host’s underwear drawer and private porn collection, causing no small amount of embarrassment and a noticeable drop in social invitations.

But Ted’s insatiable curiosity, to uncover whatever people seemed to be hiding behind a veneer of perfect lives, lived behind stucco walls and plastered smiles, served him well in a lifelong career in law enforcement, first as a homicide detective with the Los Angeles Police Department and then as a highly-paid private investigator to Hollywood’s glitterati.

According to Ted,

“You’d be surprised how many clues people leave behind. The first place to look is the garbage, always the garbage.”

“You can tell a lot by someone from their garbage. What they eat, who they’ve met, everything.”

“I’ve done cases where the person I was hired to follow had thrown their tax returns, receipts, invoices, hotel bills, everything, into a black plastic bag, zip tied it and just tossed it in the trash.”

“I mean, how much does a paper shredder cost?”

And whether it’s coffee rinds or orange peels, the devil is always in the details, according to Ted.

But because Ted has been looking for patterns since he was a child, he can offer his clients insight into their investigations from seemingly disparate and disconnected pieces of information,

“It’s all about attention to detail and it makes all the difference.”

And like private investigators, a slew of companies are now offering hard-pressed hedge fund managers, desperate to eke out what remaining alpha exists in the market, an opportunity to profit from that difference.

There’s No Alternative To Data

On the last Sunday of April in 2019, a little-known company called Quandl, which tracks the flight details of private jets, alerted its clients, mostly hedge funds that a nondescript Gulfstream V belonging to Occidental Petroleum, had touched down at the airport in Omaha, Nebraska.

A delegation from the oil company had flown in for a secret meeting with the Berkshire Hathaway Chairman Warren Buffett, hoping to persuade the “Oracle of Omaha” to come to their rescue in the battle to acquire Anadarko Petroleum.

But private jets and their flight plans are no secret, and the knowledge of the Occidental Petroleum jet’s arriving in Omaha, presented a potential trading opportunity to hedge fund investors.

Quandl’s clients were not disappointed when two days later, Occidental Petroleum announced a US$10 billion investment from Buffett.

The episode was emblematic of how the embattled fund management industry of stock pickers is struggling to recapture its edge, at a time when even the most storied quantitative hedge funds like Renaissance Technologies have seen drawdowns on their holdings.

With a growing number of traditional hedge funds underperforming and seeing their assets under management dwindle, as investors have either put their money into passive funds or computer-driven quantitative strategies, managers are turning to what has been called “alternative data.”

Traditional stock pickers are adopting some of the data-mining techniques first pioneered by their quant rivals and are investing heavily in programmers and data scientists, in the hope that a hybrid approach — combining the judgment of an experienced stock picker with insights from big data will provide them a new lease of life.

And while traditionally, fund managers have formed two distinct camps — fundamental traders who trade off macroeconomic factors and quantitative traders who look to gain an edge by using data and analytics, increasingly, both camps are converging to gain an edge.

To be sure, investors have always leveraged new technology for their advantage.

Venetian traders would use telescopes to inspect the flags of incoming ships to derive clues about their impending cargoes to buy and and sell commodities accordingly.

And Nathan Rothschild allegedly knew about the outcome of the Battle of Waterloo a full day before the Queen, allowing him to speculate on the stock exchange ahead of the market.

Jack Treynor, former editor of the CFA Institute’s Financial Analysis Journal once noted,

“You may not get rich by using all the available information, but you surely will become poor if you don’t.”

And an unending torrent of data and the difficulty of trying to make sense of all of it has fueled a revolution in the investment industry powered by advancements in artificial intelligence and big data analytics.

Nowhere is this more apparent than when it comes to cryptocurrency trading.

Cryptic Data for Cryptocurrencies

Whereas a purely quantitative approach was possible in the last two years of cryptocurrency trading, increased retail and institutional investor participation has dramatically altered the factors that cryptocurrency traders have had to program into their trading algorithms.

The integration of “alternative data,” which can range from scraping the internet for reviews, to social media chatter and web traffic can all play a significant role in volatility and volumes.

In the crypto quantitative circles, there has also been a growing appreciation for natural language processing tools, which continuously crunch millions of sources ranging from crypto Twitter to blockchain analysis, to news stories and regulatory and court filings.

And while macro factors in trading major cryptocurrencies such as Bitcoin, could be comfortably ignored prior to the coronavirus pandemic, that is no longer the case.

High profile macro investors such as Paul Tudor Jones publicly declaring their allocation of investment into Bitcoin, means that Bitcoin traders can ignore macro factors at their own peril.

And with unprecedented central bank monetary and fiscal policy measures to shore up a global economy wrecked by the coronavirus pandemic, more investors are wondering if their wealth wouldn’t be safer in gold and Bitcoin, meaning that macro matters more than ever before in Bitcoin trading.

But just like everything else in life, data, is what one makes of it.

Privately some managers lament that they have been left somewhat disappointed by their investments in highly paid programmers and data scientists — not in their efficacy, but in their ability to take advantage of trading signals for any extended period of time.

With more powerful data tools than ever before, even when data crunchers discover something they can exploit, the profitability is quickly eroded by other funds with the same datasets pouncing.

And even when the data is processed, the challenge comes from distilling that data into actionable trading strategies.

To that end, perhaps the traditional markets can take a leaf out of the playbook of the cryptocurrency markets — most data out there is either poison or garbage.

Very few things move the cryptocurrency markets — manipulation is one, and increasingly for Bitcoin at least, macro is the other.

With so much signal noise, there are very few significant markers in the cryptocurrency markets that actually have any direct bearing on price, which is why quantitative traders have been using deep learning tools for some time to filter out signal noise.

And given the relatively unregulated (and I use this term very loosely) nature of cryptocurrency trading, trading advantages can generate returns for longer periods than in traditional financial markets, where scores of data scientists and analysts are scouring every iota of data and rifling through the couch cushions to scrape out whatever loose change of alpha remains.

To understand just how “alternative” the data in cryptocurrency trading can get, consider that in recent times various data points from Glassnode, CryptoQuant and ByteTree all indicated Bitcoin miners selling Bitcoin en masse and contributing to steady selling pressure.

According to researchers from Glassnode, the largest inflow of Bitcoin to cryptocurrency exchanges was observed on June 24, 2020 — typically a precursor to large selling pressure.

And the data did not disappoint.

Bitcoin fell from about US$9,600 to as low as US$9,000 at one stage in the subsequent 24 hours — meaning that a trader who had shorted Bitcoin with a healthy amount of leverage would have made a tidy sum.

And just over a month ago, data from CryptoQuant saw the largest amount of Bitcoin flowing out from cryptocurrency exchanges, typically a sign that Bitcoin miners were holding onto their product, in anticipation of higher prices in the future.

True to form, that was also the month that Bitcoin tested US$10,300, its highest level since February this year.

Imagine trying to parse through that data and finding an actionable trading strategy without the help of quantitative tools and experience trading the cryptocurrency markets.

With more data to process than ever before, ultimately what makes a successful and consistent trading strategy is not so much data analytics tools, but the ability to synthesize data into concrete courses of action.

And that is ultimately the same challenge faced by stock pickers as well.

In an environment of unprecedented central bank monetary and fiscal stimulus, money managers, more than ever, need to rely on an entire suite of tools, whether alternative, quantitative or intuitive.

And as some of the world’s best private investigators will attest, what looks like random pieces of information to the untrained eye, are actually all clues that can help you decipher the hidden patterns in the data and from those patterns, profit.