Don’t look where everyone is looking, from Burgess Humbert

March 27, 2015 |

When planning a research agenda, I believe that it is fruitful to start with a modicum of thought about the ways in which markets try to misdirect our focus.

In extremis everything available to the modern trader that is supplied to him by the providers of the market infrastructure ( exchanges, banks, government et. al.) provides a picture or 'information/data' that in some way attempts to generate 'business' for the provider ( commissions, taxes, subservient behaviour etc. )

Whilst the above borders on tin hat conspiracy ( after all, we need SOMETHING to analyse! ) it makes one think about what factors affect prices that are not readily available.

Some are:

# True bid/ask volume and depth. Nowadays, this information as shown on DMA platforms arguably does not represent the intentions of the buyers and sellers in the market given all the different order types & not to mention HFT. So, one should research into whether magnitudes and changes in it have anything to do with future price changes in and of themselves.

# The price formation process. One has seen live trading evidence that there are very high levels of mathematics- applied to reasonably high frequency data ( not so high frequency that latency or hardware is the true 'edge') - that there exist relationships between numbers that are very predictable for short holding periods ( interestingly this type of predictive analysis descends into randomness with holding periods longer than about 36 hours - it may be of interest that the best trading firm since we left behind the primordial slime trades within this time frame) So, at the meta-level is there a price formation process that ( whilst not necessarily available to all, might in some way reveal itself )

# Changes in regime. Whilst not wanting to enter into a discussion here about things like hidden Markov switching et. al. It is very interesting to consider how, when & why the markets shift regime. This may be from Trending to non trending, from relatively low to relatively high levels of Volatility.

# Markets' varying responses to the same stimuli. An example shall suffice -yesterday morning, one was fortunate to sell GBP USD at a very good level based on one of my approaches. The market declined sharply subsequently. The same trade idea applied later in the day would have been the exact wrong thing to do. Now, I know why in terms of my trading approach but the bigger question is still there - it's all just data isn't it. Of all the four points listed in this post, this last is most reasonably addressed by an understanding of non linearity- but that for another day.

Just as I concluded the above it strikes me that perhaps 'comparative advantage' can be made to work in trading - I am certain that I have a 'closed form' answer to research that others have spent two decades on and I am sure others might look at some of my strategies' short comings and have improvements that would be very helpful to assisting one become a Rothschild.

Good day ahead all.

Ed Stewart replies:

Great post.

My working hypothesis is that markets "advertise" to draw in the maximum amount of resources to the system. Price action and "fast action" seems to entice the basic urges the way i fishing lure does for the fish. Much of the "stuff" out there, is to draw in new participants and their $$, like blinking lights at the casino. And it is not a "conspiracy" but rather spontaneous order of the market do to competing profit incentives.

Of course the above is not my idea, i learned it (expressed much more eloquently and accurately) from Victor's books. It is only "mine" in the sense that the more I trade, the more i see the applications, and the more certain I am that it is true. I would say it took me about 5 years to get the point, so I am a slow learner.

Mike Caro the poker author had an article recently that said something like "you should be happy when your opponent draws out on you". ( http://www.poker1.com/archives/12809 ) Of course, because it keeps them playing and taking shots. Same is true with trading, i think.

What the market does is advertise the "drawing out" situations and then entice people to make those plays, which are not percentage plays. Anyone who looks at a chart naturally picks out those "drawing out" situations and says, "I would have bought here and sold here" etc.

Even investors, traders, or hedgers who do not officially use charts or market-based signals can still be influenced by them, because they have psychological impact or pull that does not require conscious articulation.

Jim Sogi adds:

Big orders used to mean something in the depth, but deception is the rule of the day now. Some of the traders or market participants at CME get info on who is making the orders. That might mean something in a deceptive environment. When also seems to be important. How long seems to be a regulatory and legal grey area now as well.

Gary Phillips writes:

Deception is nothing new, and the market's micro-structure is nothing new. back in the pit, it was always my belief that the market would trade to size. floor traders would inevitably be lured into stepping in front of large orders, only to race one another as their lean got hit/taken and disappeared. the exchanges have since struck a faustian bargain to provide the same benefits to hfts, (including front-running, etc.) in exchange for the provision of market liquidity; something that was once the provenance of the floor trader. the new micro-structure, doesn't appear to have changed things much. back in the day, when we wanted to "pull liquidity" we simply put our hands down, and like the hfts, we would do the same things, at the same time.

Hernan Avella adds:

Gary observes: "and the market's micro-structure is nothing new"

When talking about the stock market, this is certainly not true. Here's a good reference Fox, Glosten,Rauterberg: The New Stock Market . When taking about futures markets, the differences are also profound. Electronic trading is a continuous auction that features an order book, because of this you add another dimension to the game: the priority in the book. You also have different types of orders that affect liquidity and risk transfer. The meta-principles of deception and whatnot might be the same in life, pits or servers, however, an observation relevant to practice has to take into account the enabling mechanisms . Hernan.

Paolo Pezzutti argues:

Overnight when trading is lighter the book may provide some " true" indications (although this should be tested…). Algos are less active, hft is not there and the slow trading pace sometimes looks more driven by arbitrageurs more than anything else. This might give an edge in an environment that has different characteristics and competitors than during regular hours.

Comments

1 Comment so far

G man on March 27, 2015 4:48 pm
Your point 2 is addressable by applying Lyapunov’s exponent

Recent Posts

List of Authors

Mar

27