Algo Trading - The Martingale Debacle of 2024
ฝัง
- เผยแพร่เมื่อ 28 ก.ย. 2024
- Find out why traders like Martingale position sizing, and why it will eventually spell disaster for algo traders...
Free algo strategies and other helpful trading info: kjtradingsyste...
#algotrading #algorithmictrading #futurestrading #Automatedtrading #kevintrades
Your Trading Guide: Kevin Davey has been algo trading futures for over 30 years, and is currently a full time trader. He is a 3 time international futures trading contest trophy winner and the author of 5 best selling trading books.
kjtradingsyste...
Kevin can be reached via e-mail at kdavey@kjtradingsystems.com.
kjtradingsyste...
These systems look appealing to a lot of beginners for the obvious reasons. I just can't have an algo with bottomless risk in my portfolio lol. Thanks for the update.
Many people fall for that first equity curve, lots of hidden risk.
Three aspects not covered in this video:
1. It doesn't have to be a 1:1 ratio - and doesn't have to be a doubling after every loss. If your R:R is 2:1, anything beyond 1.5 should be a good Martingale multiplier.
2. Using a setup that has a real edge, with a technical trigger and a technical stop loss instead of one based off of your personal risk tolerance, makes the whole thing immune to changes in market sentiment. If the market sentiment changes, the triggers align with it... e.g., in the example you gave, at some point our triggers would have been to the short side.
3. Two of the metrics you can extract from backtesting that can help with a Martingale approach are:
A. Maximum number of losses in a row
B. Average number of losses in a row
If your stats provide a small enough value for these two, you can define a maximum number of doubling-downs. Beyond that number, take the losses and either cut the cord or go back to the base leverage.
Thanks for the comments, I appreciate it. Good food for thought for anyone thinking about Martingale. My intent with the video was really to show how people blew out in July. You have put forth some ideas that people could test, so thank you!
A martingale approach is inherently a bet on mean reversion so the likelihood of getting that short signal is low and if you did then do you go back to your original $100 bet because the signal is different? Now you're really in a pickle if you have a new signal which theoretically would reset your martingale approach.
2 sigma events happen way more frequently than '2 sigma', just ask LTCM
@@AlgoMarketTheory - Thanks for the comment, good point you make!
@@AlgoMarketTheory Averaging down (dollar cost averaging or any other kind) is a bet on mean reversion. Martingale has nothing to do with averaging down and nothing to do with mean reversion. For some reason people in the MT4 community named their endless averaging down scheme "Martingale" and since then it stuck. Google: Martingale.
Great stuff Kevin!
Thanks! - Hope all is well for you Gregory!
Yes, noticed that on my demo accounts. The usual setup is averaging down when the price goes against you. The problem is account destruction if the trend continues in the wrong direction. The EAs are marketed showing good gains and low drawdown. And then disaster. A better idea is one trade at a time in the trend direction or more specifically buys when buyers > sellers and vice versa.
Thanks for the comment, glad it was demo accounts for you!
Great video Kevin. I tried martingale system before in back testing and SIM for many months. I loss with it. Not for me.
Sorry to hear that!
@AlgoTradingWithKevinDavey what's up Kevin. It's me goodoboy, hard head boy 💆♂️
Great stuff as always. Keep it up.
Thanks, will do!
Kevin, you can only give smart advice to smart people. Fools will continue doing stupid things, like martingale.
I think sometimes people might be smart, but they do not know all the risks involved...
You don't even need martingale, if you average down or double down if trade goes against you constantly, you will have the similar equity curve and the wipe out result.
Yes, eventually you run out of money!
What are some better methods to manage loosing positions without too much loss?
There is Simi 1,1,2,2,3,3 or step lots 1,2,3,4
Or hedging after x trades but the secret is finding the sauce where you make more money than you loose. Less than 1% can do this, it’s a tough gig
Very hard when there’s spread, slippage, sharp moves, small leverage
Algo trader, 4 years in, still not profitable. 😢
In a few weeks I'll be doing some free webinars, and one will be on alternative Martingale. If you are on my e-mail list, you will be invited.
What's the saying - the market can be irrational longer than youm can be solvent...something like that..
And in this case, it really wasn't even irrational, it just pullback from the recent uptrend. That is the scary part - many people get obliterated by a move that was not super huge.
I knew about this strategy for a long time. Also, I remember that was a statement that this strategy would work if you have sort of "unlimited resources. "
I had a thought while watching this video. This system needs some "circuit breaker." Something like 100 SMA + stop-loss.
That circuit breaker will likely make it a long term loser...
I dont believe anyone actually trades a Martingale approach in reality... you cant read anything on it without also finding out its basically impossible in practise..
Send me an email at kdavey@kjtradingsystems.com and I'll direct you to online reviews from people who got blown out (5 and 6 figure losses in a day or two) using Martingale type approach. So while it would be nice to hear no one does that, the reality is many people did.
I have a question regarding a simple trend following system. Lets say I want to trade futures and go long when the 50 day high is broken and go short when the 50 day low is broken. Do I still need a stop loss even if I have a 50 day low exit?
Not necessarily, but you could have quite a big loss as a long trade went from entry at 50 day high to a reverse to short at 50 day low...
Do you ever trade forex? (new here)
A little bit, but mostly futures and for currencies - primarily currency futures instead of forex.
Hi Kevin, what about "reversed martingale" ? exemple: I start with 1 lot, if i win i increase my lot with what i won. if i lose I start all over with 1 lot
yes, aka anti-martingale. If you are on my e-mail list, you will be invited to webinar on the topic later this month.
@@AlgoTradingWithKevinDavey I think I am, I will check. do you think dollar cost averaging on indices has similarities with martingale ? I think it relies on same psychological biases
@@divayht They both depend on the price recovering, which may take a long time.
I have traded the Martingale strategy before many times. It wins all the time in instruments that have a lot of volatility (move up and down) so you bet it will return to your original entry price. It wont work in instruments that may stay in one direction.
Thanks for the comment. I'd worry eventually there will be a time where the price will not return to your original entry price.
Hi kevin, thanks for making this video! I remember sending u an email regarding this, there are ways that might possibly make the system profitable, like adding variable distances, having a lower martingale multiple etc, implementing limits and sensible stops, but it seems these features are overfitting to force a perfect equity curve, i really like ur approach so far, keeping entry, exit and sizing simple and diversification to navigate the unknown future. Appreciate this! :)
Thanks for the comment. Yes, there are things you could do to make it better, and maybe I will make video on this. The key though is like you said - make it so that it wasn't just curvefitting.
Martingale or "with infinite loss aversion comes infinite margin requirements".
Very true!
Great video. I used to follow a trader who leverages up when falling below the high water mark by a certain percent, and I now realize how risky that is. I also found your discussion with Dr. Tom Starke and the Martingale paper you linked to very insightful!
Thanks, I appreciate the comment. Glad you like conversation with Dr. Tom - he is a great guy!
Like in any trading approach the question is could you manage to make more than you loose 🙃
Good point, thanks!
Thanks for the information. I appreciated it
Glad you liked it!
Personally, I would only couple anti-Martingale with trend strategies, which usually have 35-40% profitable trades. If I ever did Martingale, I would use it only as a scale trading strategy where I've calculated and know the maximum loss.
I'll probably follow up with a anti-Martingale video soon.
@@AlgoTradingWithKevinDavey what is it Anti martingale?
@@iameladlevi - Position size doubles after each win, when you lose you go back to 1. I will be doing a free webinar in next few weeks, if you are on my e-mail list you will get info to register...
I like the fact that you commented on how there is a realistic limit to the amount you can actually trade in terms of numbers of contracts. Good video in general!
Thanks, that is an important point. It is why Vegas has table limits, too!
@@AlgoTradingWithKevinDavey Have you ever looked at reverse Martingale/anti-Martingale?
@@covingtoncreek - next video!
thank you as always Kevin have a great weekend 😀
Thanks, you too!