Traders' Roundtable

I have seen claims, I forget where, that the popular bet size variation methods that we like to call anti-Martingale (reduce size on loss, increase on win) are really Martingale methods in disguise.

Can someone explain this or provide a pointer to another resource describing this?

--si

You can perform experiments yourself using, for example, Microsoft Excel. Set up sequences of coin-flips (or dice-rolls, or stock trades, or ...) and then apply Martingale betting to them. Then apply anti-Martingale betting to the same sequences. What do you see? How do the results change, and how do your conclusions change, as you vary the betting amounts? How do they change as you vary the profitability of the underlying trades/flips/rolls?

Since you asked about resources, the classic book on the subject is one that Professor Turtle, William Eckhardt, referred to repeatedly during the turtle class (according to R.S.):
http://www.amazon.com/exec/obidos/tg/de ... ingblox-20

Another excellent resource is the website bjmath.com , run by our second cousins of speculation, blackjack gamblers.

Thanks Ted for the response. The resources get filed away into my ever-increasing ToRead list.

I am sorry my post was poorly worded but I understand the issue well enough to be able to discriminate between the two. Essentially doubling when down, means that increasing amounts of capital are chasing a very small amount of profit. So it is suitable only if 1. the losing streak is guaranteed to reverse (Stocks can go bankrupt, commodities can be rangebound for ever), 2. one has the patience to wait until then and 3. one has access to infinite capital to wait it out. (LTCM)

My original question is about a specific comment saying that M.. and anti-M.. sizing both are specific instances of more general and are therefore not significantly different from each other. This struck me as being a rather strong claim, unfortunately I am not able to find this statement, the author or the context by a simple search on various fora. Possibly I misunderstood the assertion when I originally read it.

--si

Hmm, sounds like something that HarryTrader on the EliteTrader forum might have written.

Something like:

http://www.elitetrader.com/vb/showthrea ... post482840

http://www.elitetrader.com/vb/showthrea ... post481201

I'm not sure if it's me but I'm just not smart enough to figure out what he's saying half the time.

- Forum Mgmnt

Forum Mgmnt, Thanks. Those are the posts I refer to. I was beginning to wonder if it was all a hallucination.

Now on to making sense of it.

--si.

si wrote:Forum Mgmnt, Thanks. Those are the posts I refer to. I was beginning to wonder if it was all a hallucination.

I think the hallucinations are coming from a different direction. I would ask Mr. HarryTrader just how martingale and anti-martingale are the same mathematically. It's a little like saying X^0.5 is like X^2 ... in both cases X is raised to the Nth-power. The mechanism is the same but the effects are completely different.

Kevin

Kevin,

I would ask Mr. HarryTrader just how martingale and anti-martingale are the same mathematically.

Don't bother. That guy 'Harrytrader' is just taunting other readers into a frivolous academic argument with another Quant’ or mathematician to show how smart he is.

Harrytrader seems to be referring to a mathematical stochastic process that is called martingales.

A martingale appears to be related to a 'fair game'.

A submartingale is a process that's better than fair i.e The stock market over time, and a supermartingale process is worse than fair (i.e a Casino)

How that relates to anti-martigale I am not sure but if it can't be explained simply in a few words we shouldn't encourage it from him. We don't want to turn this forum into a mathematical debating chamber (Yawn).

Cheers

Stephen

Stephen: Here here! We now return to our regularly scheduled forum on mechanical systems trading

Kevin

I read the above mentioned posts from the Elite Trader board. It’s ironic that he mentions coin tossing- because I sure can't make heads or tails of any of this.

I notice that many of those who assert lots of claims about mathematics often tend to have a flimsy understanding of their assertions.

It reminds me of a post c.f. made about "quasi-experts."

Harrytrader's messages are very confusing, but the subject matter isn't that complicated. Check out a more readable source if you are interested in what a Martingale is..

http://en.wikipedia.org/wiki/Martingale

Basically if some security's daily close is a Martingale, then our best guess of tomorrow's close is today's close. There is also a discussion of sub/super martingales and the application to betting at the bottom of the page...