The Volume Clock

That is the title of a new paper by David Easley, Marcos M. Lopez de Prado, and Maureen O’Hara:

Abstract:
Over the last two centuries, technological advantages have allowed some traders to be faster than others. We argue that, contrary to popular perception, speed is not the defining characteristic that sets High Frequency Trading (HFT) apart. HFT is the natural evolution of a new trading paradigm that is characterized by strategic decisions made in a volume-clock metric. Even if the speed advantage disappears, HFT will evolve to continue exploiting Low Frequency Trading’s (LFT) structural weaknesses. However, LFT practitioners are not defenseless against HFT players, and we offer options that can help them survive and adapt to this new environment.

The paper has many interesting bits, such as this:

Databases with trillions of observations are now commonplace in financial firms. Machine learning methods, such as Nearest Neighbor or Multivariate Embedding algorithms search for patterns within a library of recorded events. This ability to process and learn from what is known as “big data” only reinforces the advantages of HFT’s “event-time” paradigm, very much like how “Deep Blue” could assign probabilities to Kasparov’s next 20 moves, based on hundreds of thousands of past games (or more recently, why Watson could outplay his Jeopardy opponents).

The upshot is that speed makes HFTs more effective, but slowing them down won’t change their basic behavior: Strategic sequential trading in event time.

One message of the paper is that sequential strategic behavior will occur at any speed.  I liked this sentence:

As we have seen, HFT algos can easily detect when there is a human in the trading room, and take advantage.

And the ending bit is this:

There is a natural balance between HFTs and LFTs. Just as in nature the number of predators is limited by the available prey, the number of HFTs is constrained by the available LFT flows. Rather than seeking “endangered species” status for LFTs (by virtue of legislative action like a Tobin tax or speed limit), it seems more efficient and less intrusive to starve some HFTs by making LFTs smarter. Carrier pigeons or dedicated fiber optic cable notwithstanding, the market still operates to provide liquidity and price discovery – only now it does it very quickly and strategically.

Comments

I'm neither a HFT or a LFT. I'm an investor, and I rarely buy or sell anything, and I find this whole thing mildly interesting, but not at all relevant to my investment decisions.

Further, all liquidity does is provide the occasional crash for an investor to take advantage of or a run-up to attempt to avoid.

(I heard what's his face, Roger Lowenstein, say part of the cause of the housing bubble was the introduction of liquidity into the housing market.)

Well, the main thing it provides is a better price

It's interesting that Open Outcry still exists or at least is the face of the Chicago Mercantile Exchange and Wall Street.

The bulk of volume on the CME is screen based. There are still some option pits where the complexity of the option packages traded makes it useful for human to human contact, i.e. details have to be clarified and/ or negotiated. When these option packages are quoted on the screen, it's easy for an unscrupulous trader to change one important detail of the package and have a market maker quote a bad price. As a market maker, it's not worth it to quote prices for brokers who are trying to trick you. That doesn't happen in the pit because of the personal contact.

"As we have seen, HFT algos can easily detect when there is a human in the trading room, and take advantage"

Seems possible then to design a 'false-flag' set of trades to lure an unsuspecting algo into thinking it has a carbon-based chump on the line...and then unleash your own superior algo to take advantage of that.

In this way it would come to resemble chess opening books from computer programs.

The market equilibrium would be in effect a kind of evolutionary stable strategy. Literally, it would begin to resemble 'deception' behavior in animal species who evolve strategies to both "cheat" and "spot the cheaters". There will be a short period of success for any algo that innovates a new way to "cheat" before the technical is discovered and the other algos learn their way around it.

But a stable strategy is not a given. It could unleash a race to the bottom of machine learning cheaters. It's creating a unstable 'market' in algorithms like this that will lead to artificial intelligence.

Databases with trillions of observations are now commonplace in financial firms. Machine learning methods, such as Nearest Neighbor or Multivariate Embedding algorithms search for patterns within a library of recorded events. This ability to process and learn from what is known as “big data” only reinforces the advantages of HFT’s “event-time” paradigm, very much like how “Deep Blue” could assign probabilities to Kasparov’s next 20 moves, based on hundreds of thousands of past games (or more recently, why Watson could outplay his Jeopardy opponents).

This sounds to me like fairly sophisticated AI. It certainly could be (is) smarter than humans are at certain kinds of investment decisions. Remind me again why we like it when robots steal factory workers' jobs but get upset when they try to replace traders?

That depends, can I get a robot trader on my side like I can a robot manufacturer?

I don't day trade. I will never day trade. If the computers are willing to day-trade against each other and provide liquidity in case I need to write a check for a $10,000 roof repair, I'm all for it.

Perhaps I'm misreading this post and abstract but it seems like fairness complaints about HFT focus on, you know, the HF part. Saying one shouldn't complain about a speed advantage because they use sophisticated algorithms along with a speed advantage is illogical. What am I missing?

Rather than seeking “endangered species” status for LFTs (by virtue of legislative action like a Tobin tax or speed limit), it seems more efficient and less intrusive to starve some HFTs by making LFTs smarter.
How would one go about making LFTs smarter? Wouldn't HFTs just update their tactics as well?

If this is what HFT profitably (i.e. earning supranormal risk-adjusted returns) do, is this not in essence contradicting weak form EMH? Analyzing trillions of past events sound very much like technical analysis to me...

If 1,000 computer algorithms collapse due to a flash crash, does it make a sound?

EMH is approximately true precisely because of these.

Aye, this is merely a (small) move towards greater efficiency by way of allowing the exploitation of more arbitrage opportunities. EMH isn't a law, it's a target.

All things being equal, higher speed trading should improve price discovery. But all things are not equal if HFT drives small investors out of the market.
If sprinting was as expensive as polo, would Usain Bolt have been discovered?

If I recall correctly, when the flash crash occurred the markets rolled back some number of trades. I'd like to know why that was done. It seems like a free insurance policy for HFT traders. If your algorithms go nuts in a big way, the market rolls it back. I think everybody who got burned in the flash crash should have stayed burned. Anything else is not a free and fair market.

Busted trades usually favor low frequency people who made a mistake. Ordinarily HFTs identify mispriced contracts in a matter of microseconds and take them out. Still want to jam people with off-market trades?

A lot of the comments are referring to the ”price discovery” – but doesn’t HFT mainly add to the efficiency of “ketchup prices” (by improving the “ketchup economics”).

Also, unless everybody with the same risk profile holds the exact same portfolio I can’t see why there even would be a theoretical “true/correct” price for any financial asset which makes it hard to see how even that limitedly defined price discovery would be of any particular benefit.

Could anyone give me a reference as to where someone explains it?

Wouldn't HFT programs all tend to make the same or very similar trades , and wouldn't this lead to market distortions ? Isn't it likely that such distortions are already happening ?

This, I think, is already the case with regular LF technical trading.

Every financial investor I know (even the day traders) are using one or another version of the trend following strategy, which reasonably leads to amplified swings in the market.

“The price of gold is rising?” – “Time to buy!” – “If there are any fundamentals to explain it? What kind of question is that? Who cares WHY it is rising. The probability that a price will rise in t+1 given that it has risen by x in the last y days/months is >50 %”

Comments for this post are closed