Hacker News Reader

It's hard to tell wash trading from legitimate trading.

Because I'm involved with a committee on financial semantics I wound up learning a bit about swap trading. For stocks if you don't like your long or short position you can buy or sell and it is done.

In the case of swaps if you don't like your position you write another swap contract that is the opposite of the one you don't like. Both are on the books. In the 2008 crisis the size of outstanding swap liabilities dwarfed the real economy, but when you added them all up they mostly canceled out, both in the aggregate and for almost all of the market participants.

Looking at a situation like that which is hard to unravel people are going to make assumptions about the motives and ethics of the participants which are not substantiated.

It will be nigh impossible to prove wash trading without the exchange being very obvious or primary trader identification.

Famously, a Mt. Gox data leak actually proved wash trading on that exchange conclusively as same trader IDs took their own orders. [1] So there definitely is proven precedent in the crypto market.

I am not saying it doesn't happen and isn't likely on unaudited/unregulated exchanges. I just want to highlight that the authors make very strong claims and alternative explanations should be explored.

1. https://dx.doi.org/10.2139/ssrn.3362153

The bigger factor in the size of the swap market dwarfing the size of the economy is due to the former being reported in terms of notional.

I will happily enter into an interest rate swap with you on SOFR^-14 and a notional of USD 1 trillion.

> but when you added them all up they mostly canceled out,

Except they don't cancel when one party goes bankrupt and a bankruptcy court decides how the assets should be divided.

I came looking for your comment. Someone applying some common sense to the “research”. Yet, the post has >400 points.

Feels like these days anybody can post a paper and get uncountable widespread with 0 backings for their research.

In nutrition this happens A LOT. Things like: “meat causes cancer in 70% of the population”. And then you read the paper and they did the study on 80 people between 60-90 years old. There’s just no scientific/statistic rigurosity.

you can make something up at the speed of thought. refuting it takes time and effort. peer-reviewing even longer.

Grid trading bots are all over the place. If you look at the crypto-trading subreddits you'll see plenty of posts about how some guy downloaded a bot and makes $300-500/day off of it. These are the new scr1pt k1dd13s; there are turn-key solutions to automate trading and anyone can use them.

It's also very disingenuous for the title to say that 70% of the volume in the top crypto exchanges is wash trading when the 70% category they define are the least popular exchanges (they rank worse than 960 on the finance section of similarweb).

Even more so because the regulated exchange and popular unregulated exchanges have 0 and mostly <20% wash trading respectively.

Their idea does show a pretty stark difference b/w "regulated" and "unregulated" exchanges, however, as would be expected if that activity is fraudulent in some way.

I'm no fan of crypto currencies (as my comment history will show), but does "different" imply "fraudulent" necessarily? or are there other factors at play? eg. easier APIs, more variance in crypto prices so that an automated trading strategy becomes potentially more lucrative?

Benford's Law is a pretty established method of detecting fraud in forensic accounting.

You don't even need to detect it. Name a single instance in human history where an unregulated and unaccountable industry didn't instantly degenerate into fraud. "2.5T" "dollars" (majority held by early insiders) is an incredible moral hazard

"I'm sure there is a lot of fraud" is very different from "This is the type of fraud that is going on and here is the evidence".

The difference matters for legal proceedings but not for much else, surely not for personal investment decisions.

Benford's Law requires that the numbers being analyzed typically occur over a wide distribution spanning orders of magnitude, which typically doesn't happen in trading (except over very long time periods).

Why would regulated and non-regulated exchanges differ in whether they follow Benford's law or not then? (As in the article)

Benford's law is extremely dubious in the field. [1]

"Abstract. Is Benford's law a good instrument to detect fraud in reports of statistical and scientific data? For a valid test, the probability of ‘false positives’ and ‘false negatives’ has to be low. However, it is very doubtful whether the Benford distribution is an appropriate tool to discriminate between manipulated and non-manipulated estimates. Further research should focus more on the validity of the test and test results should be interpreted more carefully."

1. https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1468-0475....

That is not representative of "the field". Your paper, from 2010, has 21 papers citing it on the link you provided. None support the paper conclusions. The highest cited one of those, from 2019, ([1] with 172 citations) uses Benford's Law exactly as it's commonly stated and used.

Here [2] is Google Scholar on Benford's Law. Of the 36,000 papers it pulled very, very few claim the law is not useful and valid. The vast majority (actually, every one of the first many pages) show how it's useful and demonstrate uses of it. If you want only recent papers, select from the left panel. Same result.

In fact, the first several pages of results contain many papers showing empirical validation of the usefulness of the law.

[1] https://www.degruyter.com/document/doi/10.1111/j.1468-0475.2...

[2] https://scholar.google.com/scholar?hl=en&as_sdt=0%2C14&q=ben...

> Benford's Law is a pretty established method of detecting fraud in forensic accounting

Benford law never had to deal with exchanges where any customer can write a python bot and start trading at sub-second latencies via the exchange API's.

I'd be very surprised if whatever statistical model they're relying on is in any way a match for what real (as in: legitimate orders from actual customers) trading goes on on crypto exchanges.

Python ... Subsecond latencies

Ok

>Python ... Subsecond latencies

Lol, fair enough.

But I believe my point stands even if you go to python-level latencies.

Bitcoin transactions with sub second latency

Double ok

You dont trade on the blockchain. We build software that does multi hundred millions orders per day. Millions of trades. But of course, not in python.

In Go, lol

"Subsecond" isn't much of a bar. A second is an eternity.

Benford's law does not apply here. Benford's law is designed to separate human generated numbers from truly random ones. The numbers generated by exchange trades are generated by humans choosing trade sizes. This entire methodology is a complete joke.

Only if you assume "automated" == "fraud", which is the distinction that grandparent is getting at.

Imagine you have a completely manual market which follows normal statistical distributions, including Benford's Law. Now you introduce one grid-trading bot with a large amount of capital. (A grid trading bot is basically a piece of software that automatically buys when the price falls below a certain level and then automatically sells when it goes above a certain range.) That one bot is going to make up the vast majority of transactions, because it's effectively "clamping" the price within its trading range. When random fluctuations take it below, it buys and sets a floor on the price. When random fluctuations take it above, it sells and sets a ceiling on the price. If you make the range small enough that most ordinary trades would end up occurring with the bot, it's going to take up the vast majority of volume.

There's nothing illegal about grid trading. They work to dampen random price fluctuations in a market. In exchange, they take on the risk that fundamental supply & demand might shift enough that they're left holding the bag, eg. they run out of inventory to sell and then the price jumps sharply higher, or they collect all the inventory and then the price drops.

But because they're non-human and take one side or another of most trades, they are going to account for a disproportionate amount of volume. This isn't fraud, it's that you don't understand the structure of the market.

> Automated trading strategies (e.g., "grid trading") are really popular

Why? After buying and selling side fees, is it easy to make a profit in an automated way with crypto?

I don't know, but it seems to be. In general though, if there are easy profits to make it is because someone on the other side is willing to lose "a small amount of money" for a long time knowing that when things change they will make a ton of money fast, on the backs of all the people they lost to.

For starters, if you can make money in automated trading, why would you not make all the possible money yourself instead of telling other people (or writing the software) and splitting the profit? This question is one I ask anytime someone mentions easy trading, and because I've never heard a good answer I stay away.

To do better than average is possible, but it requires a lot of deep study.

Because the retail market is big. If you personally buy or sell, you don't care for the "perfect" price.

I wrote a trading bot, it worked well for a while and made money until Mt Gox shut down. I open sourced it because I thought it might be interesting for others to see.

I updated it for Bittrex, but didn't have enough capital to actually make it work.

Now everything has gone up so much, and I'm bearish on crypto in general, so I don't want to have any money invested in it.

That's why mine is available, and I'm not currently trying to keep it running.

First, most good trading, crypto or not, is systematic, which is automated.

Then, from what people have told me, crypto is a highly inefficient market so huge of potential for market-making strategies. Tons of arbitrage opportunities between the various venues, centralized and decentralized, mostly uninformed non-professionals so non-toxic flow, huge spreads, etc. It's like the traditional tradi'g world but 30 years in the past.

"Grid trading" is apparently just a weird term for placing orders proactively on multiple levels to obtain a better queue position, useful for increasing edge capture on exchanges with price-time matching. There are a lot more sophisticated techniques you could transfer from the world of traditional electronic trading. Of they're successful on much more competitive markets, no reason they wouldn't work on crypto.

If I had vested interest in crypto, of course I will dismiss this paper as bogus