Please No! (Martingale)

Last time we dug into “Anti Martingale” – a method of bet size adjustment to capitalise on your winning streaks. Today I want to look at Martingale.

“But Mark, you are mad, martingale is like playing Russian roulette”

Yes, I agree. It’s deadly and IS NOT ADVISED for reasons we will soon see. But, let’s stay open-minded.

We never discover new things unless we have the ability to assess even the most absurd things objectively.

What is martingale?

Martingale is simply doubling your bet size (risk per trade) each time until you win.

Eg:

Let’s say you have an RvR of 1 and a 50% chance of winning a trade setup and you risk £100 initially.

Trade 1: Risk £100. Result = Loss – £100
Trade 2: Risk £200. Result = Loss – £200 (-£300 total)
Trade 3: Risk £400. Result = Loss – £400 (-£700 total)
Trade 4: Risk £800. Result = Win + £800 (+£100 total)

You can see how we double the trade risk after each loser until we get a winner and then start again.

Mathematically you CANT LOSE!

But there’s one small problem…

That only applies if you have an infinite bankroll. Which no one does. So in actual fact, it’s deadly. Doubling your risk on each trade gets bum-clenching pretty quickly.

You start by risking £100, after ten losers you are now risking over £50k per trade. A few more losers and you are in the millions. Madness.

Alright, so I think we are all agreed that on the face of it, martingale is a surefire way to bankruptcy. But let’s keep being open-minded…

How COULD it work?

(I like to force myself to solve ridiculous problems.)

EG:

“You are forced to use a martingale strategy in your trading, how would you do it”

I think we need to look at this differently than just doubling after each losing trade.

Ten losing trades in a row, when you are not in tune and potentially on tilt, isn’t such a long shot. We need to strip out the human element and just look at the market.

Ok, so unlike say a roulette wheel that has no memory. (The chance of red or black is always 50/50 on each spin regardless of the prior result)

The market we could argue does have memory.

The mean reversion effect is real.

If the broader market is down for 5 days in a row, each additional day makes the rubber band more stretched and the market more likely to bounce for a day.

(I appreciate this is really a back-of-envelope theory, but it’s a good place to start)

So if you HAD to use martingale, perhaps you could look back at historic market streaks.

Start to fade a daily move after say 5 losing days in a row.

Go long on day 6 and double your position size until the market finally bounces.

Now of course the problem here is one red day is not like the other. DOW down 800pts is a red day, just like DOW down 4pts is.

You’d have to find a way to normalise that somehow. (An options strategy perhaps).

I dunno.

I just know it’s not for me.

I’ve seen a lot of market moves in my 20+ years at the screens and supposed once-in-a-lifetime events seem to happen every few years.

Using any method that has even a slim chance of ruin seems foolish to me.

One unusual “once in a lifetime” event and it’s game over.

I love this business too much to be knocked out by a market anomaly I’ve seen many times before!

Today, DON’T think about how martingale could be used in your trading. Focus on your risk management instead!