Drawdown and holding time

Cheng · Post by **Cheng** » Mon Sep 09, 2013 10:50 am

FDax: +1 for the last paragraph !

goldorak · Post by **goldorak** » Mon Sep 09, 2013 11:04 am

@Tradenator: I don't think there is a difference between "a position", "a trading strategy" or "a fund". It all comes down to the same thing.

@FDAX: I could not agree more with you with respect to the volatility/drawdown idea. From my point of view, the problem is to use the expected value of something as a measure of risk. People can use whatever "something" they want, but relying on the expected value of that something is a sure way to miss your objectives by all metrics.

My own view is that there is a fundamental difference between "fluctuations" and "risk". Fluctuations are a temporary loss of capital. Risk is a permanent loss of capital. The risk associated with volatility comes down to the investor not choosing the right level of fluctuations he can hold, resulting in premature "cut your losses" or "desinvestment" decisions. As such, the risk is not the volatility itself, but a bad decision made on random noise. And of course, the higher the noise, the more difficult it is to take a rational decision!

@euroskeptic: sorry for the hijack

euroskeptic · Post by **euroskeptic** » Wed Sep 11, 2013 2:14 pm

@#goldorak,

Please do! I am learning something from you...

As a side but related issue, are there any recent studies that test the efficient market hypothesis? I am aware some researchers claim that structural market changes make then inefficient. Has that been tested for US markets? What about commodity and forex markets?

Cheng · Post by **Cheng** » Wed Sep 11, 2013 4:58 pm

Google for Buffet's paper about the Superinvestors of Graham-and-Doddsville. It is not very scientific but gives a good idea on how "efficient" some marktes are (and have been for some time).

purbani · Post by **purbani** » Thu Sep 12, 2013 1:47 am

A basic Excel and VBA implementation of the Lopez de Prado method for the Gaussian case only here

FDAXHunter · Post by **FDAXHunter** » Thu Sep 12, 2013 2:14 am

goldorak: . Fluctuations are a temporary loss of capital. Risk is a permanent loss of capital.

Unfortunately I don't have a way of differentiating between temporary and permanent ex-ante. So a loss is a loss to me. Maybe I'm old-school in the sense that I value mark-to-market more than mark-to-whatever, but if you have a way of differentiating your temporary from your permanent losses, then please teach me! I promise I learn quickly.

I don't see why the expected value is not useful. Surely an expected value is still strictly better than no value at all. Furthermore, it goes without saying that in most cases, if you can calculate an expected value, you're usually in position to obtain some sort of dispersion of the variable as well, which again, may not be the information you would ultimately want to have (that would be knowledge of the worst-case outcome). Nevertheless, from a practical point of view, this shouldn't be too big of a concern.

AndyM · Post by **AndyM** » Thu Sep 12, 2013 5:56 pm

I agree with both of you to a degree; on the one hand, money is money, there are no style points for losing it artfully. On the other hand, it's sometimes clear enough that you're the wrong way round facing a stop loss cascade, and then you frame the question not whether you should get out, but whether you can hold on. This is from the perspective of someone who trades a lot of STIRS, where you're trading defined outcomes (modulo the basis), rather than something where value is more nebulous, be it FI, equities, FX, commodities, whatever.

example: I am long Eurodollars. Market gets a weed up its ass that the Fed is going to hike 50bp at the next meeting...loss:temporary (fluctuations). Fed does hike 50bp at next meeting: loss permanent (risk).

I only think this way about STIRs (and sports betting); in all other domains, fluctuations = risk, to my mind; even if I have strong ideas, I respect Keynes' dictum.

(obviously, there's more to it, since CBs can be led by market pricing; there are no categorical answers, it's just a predisposition / mode of thinking).

Cheng · Post by **Cheng** » Fri Sep 13, 2013 9:04 am

I only think this way about STIRs (and sports betting)

I think you made a pretty important point here. Whether you act on the premise of sports betting or card counting (ie mean reversion) should make a difference imo. I personally follow a card counting approach, so I don't care too much about short-term fluctuations as long as I have good enough reasons to believe that mean reversion kicks in at some point (the obvious risk is that it doesn't, resulting in permanent loss of capital or at least missed better opportunities, think MSFT for example).