How long did you take to make money

[dakr]

Hi all,

Just joined the forum today after someone recommended it.

I've been working on algorithmic trading for the last 3 years. I have a full time job, so I do this just about every night and every weekend. The first 1.5 years was just on data sourcing and cleansing. Then generating signals and testing various strategies.

So far my result has been zilch. I haven't found not one strategy that is worth taking live. Now feeling defeated and dejected after the last idea turned out to be a dud as well.

Just wondering what other people's experiences are. How long did it take you to become profitable. Are there any pitfalls or secret sauce to become successful?

Thanks.

 

[Real Money]

So where are you sourcing the data and how do you clean it? Real time? What are the related expenses running?

 

[tommcginnis]

If you're trying to write algos before you know how to trade, it won't work.
Figure out a dynamic market first, then reduce your routine to rules, then code those rules.
THEN test, backtest, retest, rewrite, tweak, geek, and freak.
In one word: SCREENTIME.

 

[gaussian]

I have a full time job, so I do this just about every night and every weekend. The first 1.5 years was just on data sourcing and cleansing. Then generating signals and testing various strategies.

Relax. If you're spending more than 30 minutes to an hour a night finding setups you're wasting time.

No matter how smart you think you are, no matter what technology you throw at it, you will discover spurious correlations. Strategy hunting by brute force is awful and a great way to go broke. You must have a hypothesis, collect data, test it, and repeat.

Again, it bears repeating you are wasting your time. Spend exactly 30 minutes and no more than 1 hour looking at charts. Pick a trade and take it. If you cannot trade a system manually you cannot program one. You will lose money - this is tuition. Stop wasting time programming and get money into the market. It wasn't until I realized this that I made money.

I spent many, many months hunting for automated strategies - a complete waste of time. So many man-hours wasted on chasing the ephemeral "golden goose" because all the books I read said I just needed to program some bullshit and I'd be driving ferraris. They're all full of shit. Every one of them. The famous ones (many of them sponsor this forum - specifically the algorithmic trading section) have hero-stories where they came back from nothing and became wealthy. What they don't mention is they did this in one of the longest bull markets in history. They also don't mention how deeply capitalized they are, what execution broker they use, etc. Everyone is out to mislead you and many times part you from your precious trading money with promises of boundless riches if you just do this one thing. If you have to ask every single trader writing books is a liar unless they say they are terrible at trading PERIOD.

I didn't start making money until I traded options manually and just got in to the market. I spend exactly 1 hour, many times less, looking for interesting setups with filters, software, etc on the 1 day, 1 week, and 1 month charts. I augment myself with some software to help me keep abreast of news but that is it. I place trades myself because this is what works for me. Now, you need to find what works for you. Just don't waste time looking for the perfect trade. Take the one that's a little sketch - it's the only way you'll learn.

 

[Peter8519]

No algo to share. I can do programming but I have to admit I can never program to tame the market beast. So, my next best approach is to follow the trend. Using DJI and CBOE as examples. 2017 is perfect for trend following strategy but not in 2018 and 2019. But it's still OK if we reduce the position size and set wider stops in 2018 and 2019. Those stocks with less severe correction than the DJI in Oct to Dec 2018, 60% of them perform better than DJI in Feb 2019. So, narrow the target and improve the odds.

 

[Laissez Faire]

Programming can be very useful for many things. I certainly wish I was proficient at coding. I've had programmers help me create the tools I use. But programming is no substitute for market understanding. It may not relate to you personally, but I feel many people are trying to solve the puzzle of the market by applying programming without first understanding the market.

Market understanding first. Then (if at all) coding.

Big picture - 3 years part time is not that much, IMO. At least if you started with a clean slate.

Maybe work out a plan of what you really want to do in order to narrow down your focus?

Do you want to day trade? How many trades per day, i.e., frequency? Fully automated? One instrument only? What asset class? A large basket of stocks/commodities? Longer time frame trading on multiple instruments? Options?

There's really so much you can do in the markets and so many paths you can walk down where each can make you money, but you need to choose one first, IMO, and narrow things down. After that, except several more years before you're likely to get anywhere.

 

[Turveyd]

As already said.

Find a profitable simple plan thats codeable first, so chart work to find, lots of chart work, remember simple.

Im trading mainly of m1 chart Envelope 6sma 0.03%, enter with direction if up near the low, exit on envelope 6sma 0.06% or direction change,profit is the tricky part, dax and nas100 tested.

Code that, with a fixed TP say 6pts, that will get you started, improve later.

I havent automated as TP highly subjective, only my SL is coded, I’m a system programmer by trade, not in any trading related way mind.

 

[Fonz]

It took me more than 10 years to stop programming, to stop using oscillators and indicators, to mostly use naked chart and fundamentals and eventually being profitable.
I trade anything taxed as 1256 in the US (short term profits are less taxed), like futures and indices.
Nothing exciting, just patterns, momentum and events.

 

[maxinger]

It took me many many years to

- stop listening to what the financial 'experts' are talking
- decontaminate my mind from all the wrong teachings from trading coaches
- delete all the 'famous' useless worthless indicators and oscillators and whatever
- stop looking for best, most accurate financial & economic news
- stop predicting the market and stop improving my trading accuracy

Then I started to make progress.

This is my ultimate chart with no algo programming stuff.
I used to have thousands of lines and curves and sun rays and rainbow bands and stars and those things on my chart.
Now it is simply very plain.

 

[danielc1]

The years I spend on 'research' to find a working trading system are years wasted. No wait... Not wasted, it brought me to a point of so much frustration that I looked at trading in a different way. Now I know that trading results are simple an outcome of random events, with a probability in your favor. My biggest problem was that I was smart and stupid to not realise I did not understand it. I thought I did. But with a technical background I had troubles with the unknown as a variable. I was searching to get a secured way of knowing the unknown. More research and testing, and so on. Untill one day it was as clear as a sunny day. I did not understood what random means, probability and how I reacted it on trade by trade results. I have found many threads on Elite with really sharp people about coding, algorithmes and complex trading system that could cure cancer in my opinion, but has zero results in a real trading envoriment.