Benford's Law is a basic, counter-intuitive mathematical principle that has helped governments catch tax cheats and other fraud. But don't worry about any of the complicated mathematics. All that matters for the purposes of this article is that Benford's law can be used to detect if there is any human manipulation in a set of numbers. Before we get to the fraud part of the article, just a quick two sentences on Benford's law. According to Arno Berger, Benford's Law states that "in many naturally occurring collections of numbers, the leading significant digit is likely to be small." The law follows a logarithmic distribution, which in layman's terms means that if you have a random collection of numbers, the first digit of those numbers will be a "1" 30% of the time, a "2" 17% of the time, and so on, with a "9" occurring as the first digit only 5% of the time. While the law is subject to some natural preconditions and does not apply to statistics like human height, for example, it can hide in the most unusual places, like sports statistics, city populations, and more.
If any of that was unclear to you, remember that all you need to take away from that last paragraph for the purposes of this article is that Benford's law is extremely powerful in detecting fraud in a wide variety of disciplines. Learning about Benford's Law got me thinking that perhaps I could use this mathematical tool in trying to detect fraud in the cryptocurrency space, particularly, in regards to Chainlink ($LINK). $LINK has been on a tear over the last month or so, its price nearly quadrupling to $16 from last month's steady price of $4. This massive pump has led many to hop on the $LINK bandwagon, cheering on the digital asset on Twitter and sites like Publish0x.
However, some have claimed that this pump is the first phase of an inevitable pump and dump scheme orchestrated by some shady Russians behind the Chainlink project. Most notably, Zeus Capital, an asset management firm focused on alternative investments, recently came out with a pretty damning report claiming that $LINK is a fraud and its price will go to zero faster than you can say YAM. While the report scared me off from investing my modest trading budget into $LINK, I have been frustrated by the profits I have foregone — I first invested in $LINK when it was at roughly $5. And so, I set out to apply Benford's law to some of $LINK's numbers.
I didn't use any fancy tools to do my analysis, opting instead for some basic Excel programming. I started by scraping data I could find on $LINK's daily volume and I calculated the change in daily volume for six months worth of data. I chose volume as my metric because Zeus Capital claims that a majority of $LINK is held by a few of the original developers, and that those developers manipulate the volume to pump up the price of the digital asset. I then used some Excel formulas to calculate the frequency of the leading digits of daily change in volume and compared the data to Benford's logarithmic scale. And, to my surprise, the numbers matched up! The first digit was a "1" almost exactly 30% of the time, a "2" almost exactly 17% of the time, and so on.
If you think the same way I do, your initial reaction is probably, "Great! So $LINK is not a fraud after all. And there's no time like the present to invest my life savings in crypto." But after mulling things over, I realized something important that I had missed in constructing my analysis. If the developers of $LINK are aware of Benford's law — and honestly, you never know with these wily Russians (just kidding) — then they could easily avoid detection by simply making sure that their manipulations line up with the percentages necessary to satisfy the law. In other words, they could artificially make the numbers look like a natural collection of numbers by carefully calibrating their fraud to comply with Benford's Law. Thus, my findings don't necessarily mean that $LINK is not a fraud.
In the end, my crude attempt at mathematical analysis seemed to support the notion that $LINK is not a fraud. However, I do not purport to be any sort of math genius, and I could easily have overlooked something simple in setting up the conditions of my number experiment. And with so many other good crypto investments out there, it is probably most prudent to simply stay away from $LINK altogether — at least that's what I'll be doing when I trade. If there are any numberphiles out there, please let me know what you think of my experiment!
