May

20

Which are the flaws of trend following strategies? For me, the markets are homeostatic but not in a strictly way, they are like a thermostat trying to keep a room’s temperature steady. But, sometimes they can spiral into imbalance when people’s actions and beliefs feed off each other.

I’m amazed by how this dynamic of the markets has never been of particular interest in the academic world. It’s years since I don’t put in serious research given my focus on active trading, but unless something changed over this time, The classics wisdom is that markets are (let’s leave the perfectly efficient markets alone) in a perfect world supply demand state, I don’t remember having read of documented positive feedback loops etc, and overshooting and disequilibrium, which are clearly the case. I studied economics and I always found funny how some theories are.

William Huggins writes:

econ is not the right place to learn about markets because the demands of "equilibrium" require us to bend reality into the "preferred" (supposed logical) state. what most EM theorist missed was the friction in markets (adjustment time, imperfect info, etc). but momentum has been documented by financial theorists for decades. you're just digging in the wrong part of the field.

Francesco Sabella responds:

I know momentum has been documented, my point was about my economics classes, where I’ve never read about the topic, but also about how the mainstream academia is built - disequilibrium is not the classic wisdom and not even considered; you’re right that some academics have addressed momentum, but the mainstream economics is still fixated with equilibrium.

William Huggins adds:

econ likes equilibrium because its calculable, and because it approximates a physical science - if only those pesky active agents weren't so concerned about their perception of what matters instead of just obeying economists' assumptions!

in the end, its -hard- to math out the implications of imperfect, changing perceptions among a changing cast of strategically interacting characters when the targets are moving and trading may (or may not) be informed. because of these challenges, most prefer to stick to sanitized general equilibrium models, even if those are basically mathematical masturbation based on definable, stable, continuously differentiable utility functions. people usually find it expedient when modelling to pretend the net impact of all the above microstructure is "random" (its not really random, more like "too complex to model") within some defined distribution, which itself is a fudge but as Bachelier showed in 1900, might not be a bad one (and it helps avoid overfitting your forecast models).

here's are a couple of articles less than 30 years old on dynamic disequilibria (i cited Jegadeesh and Titman years ago so they were top of mind). the first talks about implications for macro, the second for asset pricing:

Towards a Dynamic Disequilibrium Theory with Randomness

Quantum Equilibrium-Disequilibrium”: Asset price dynamics, symmetry breaking, and defaults as dissipative instantons


Comments

Name

Email

Website

Speak your mind

Archives

Resources & Links

Search