Welcome to NexusFi: the best trading community on the planet, with over 150,000 members Sign Up Now for Free
Genuine reviews from real traders, not fake reviews from stealth vendors
Quality education from leading professional traders
We are a friendly, helpful, and positive community
We do not tolerate rude behavior, trolling, or vendors advertising in posts
We are here to help, just let us know what you need
You'll need to register in order to view the content of the threads and start contributing to our community. It's free for basic access, or support us by becoming an Elite Member -- see if you qualify for a discount below.
-- Big Mike, Site Administrator
(If you already have an account, login at the top of the page)
After spending quite an intensive couple of years in developing, backtesting and trading (all automated) various systems, I try to take a step back and evaluate the overall paradigm, and better define my objectives.
According to my understanding, there are 2 main schools of thought.
1. Identify resilient patterns that work in all, or many, different markets, and persist over many years. I understand this is what people like James Harry Simons and Jaffray Woodriff did (based on interviews I watched and read).
2. Don't aim at the holy grail, instead, use systems that fit a certain market regime, and for specific markets / instruments. Then, use dynamic position sizing and other ways to switch between systems in order to adapt. This kind of concept is communicated in the books of Howard Bandy and was described online in another forum by a trader called acrary.
Personally I use both systems that test well on all the data I have (and continue to perform nicely out of sample), and systems that worked only in specific periods, for example on the last 3 years.
It will be nice to stick to (1), but it is obviously easier to develop systems that are less robust as in (2). I also noticed that when fitting to specific market regimes, the yield (under the same risk profile) that can be generated is higher then when trying to find something that performed well over all of the data.
The question I struggle with: is the constant adjusting / adapting sustainable? don't you lose too much profit when moving from one system to another?
I recall from Hedge Fund Market Geniuses that Woodriff specifically mentions he became successful when he stopped developing systems that fit a single market and started to use patterns that work in all markets.
On the other hand, in changing markets, is it realistic to hope to find edges that didn't erode over 10 or 20 years?
It will be interesting to hear other opinions on this subject.
There was the move from pit to screen trading
There was the up-come of HFT
There was the slowly erosion on HFT
Today the big think is AI and the possibility to mine for 'features'
Then there is the markets itself, central banks who pumped billions of liquidity into the market, that was/is
a new phenomena, the unwinding will be an equally new phenomena, etc..
So my answer is 100% sure : no systems survives 10 or 20 years
I tend to think the same. However, when I read stuff like the quoted part below, from an interview with Jaffray Woodriff in "Hedge Fund Market Wizards "(published 2012) I am not so sure:
You might want to read Mike Harris's (Price Action Lab's) Fooled by Technical Analysis. I haven't read it yet because it is only available through his website and there are not many reviews. While the title is "Fooled by Technical Analysis" but the topics seem to go into system development. But, I have seen him present some unique insights from what his work that I have read.
I think 10 or 20 years is a very long time to obtain success from markets-- even 5 years is doing good.
The answer to your question is much to do with the non stationary nature of markets. If you can adapt to trade a single market regime well then you can make a significant outsize return even though that return won't last. On the other hand, if you want to trade the same system every year, if you want more confidence, you will need to use a longer backtest history and your returns will be lower (historically) but you might have a stronger confidence what you are doing can work going forward.
I guess it depends on what market you are trading. I think there are edges that last for 10 or maybe even 20 years. However, the markets do change structurally and also in other ways. It is reasonable that the best/biggest returning edges will be edges that are newly discovered and that the returns will decline as more traders discover them or as you suggest markets change due to natural processes.
So, the best edges today to trade probably do not even exist in the data. And, with more and more pros trading against each other, this is why I think focusing on market cognition can be just as important as focusing on system development. Market cognition refers to how markets are operating today, how they are working, how traders are trading, etc.
I think the answer is also whether or not you are okay with average returns or if you want to get into the top 10%. Right, it makes sense as you seek for higher returns that you are taking greater risk. The risk can come in many forms including various types of unknowns.