I did 2 years of back testing and 6 months of front testing 'live' before putting a strategy of my own in place. I feel like you can never be to certain. Lots of strategies will work for a window of time, but once the market environment shifts it totally changes things. You need to see if your strategy can withstand those changes.
You are using an out of date browser. It may not display this or other websites correctly.
You should upgrade or use an alternative browser.
You should upgrade or use an alternative browser.
how much live trading is relevant?
- Thread starter flatron
- Start date
livevol_ophir
ET Sponsor
Quote from TraderZones:
You WORK for them, or you OWN them? Why would an employee be a sponsor unless they had a financial interest? "Hi, I am employed by McDonald's, and I am paying a sponsorship to get everyone to eat there..."
Odd...
I work for them. You think it's odd that an employee believe in the product of their employer? I think it's odd that you think it's odd. Why would I work for them otherwise? Maybe I misunderstood?
Given that you actually have live results, it might be worthwhile to backtest your system on batches of data from similar market conditions and compare the two.
If they track fairly closely, this should give you more confidence in the robustness of your backtesting method. This would prove useful if you create a similar system or need to tweak some parameters of your current setup.
My $0.02.
If they track fairly closely, this should give you more confidence in the robustness of your backtesting method. This would prove useful if you create a similar system or need to tweak some parameters of your current setup.
My $0.02.
livevol_ophir
ET Sponsor
Quote from TheGoonior:
Given that you actually have live results, it might be worthwhile to backtest your system on batches of data from similar market conditions and compare the two.
If they track fairly closely, this should give you more confidence in the robustness of your backtesting method. This would prove useful if you create a similar system or need to tweak some parameters of your current setup.
My $0.02.
Agreed in full...
Hello Swan Noir,Quote from Swan Noir:
Rabbitone, I have been following this thread with interest and I appreciate how clearly you lay out your thoughts and practices. And kudos to Abbitia, his fine questions help illuminate the topic.
I am not sure I understand why you consider a 4% loss in a week devastating for a day trader. I currently day trade NQ and a handfull of currencies in a small account and have had weeks where a 6% loss, because it was within my plan's parameters, did not shake me. I didn't like it but it was not an albatross.
Obviously a small account generally takes a higher percentage risk than a larger one and subsequently runs a higher risk of ruin. As in most businesses being under capitalized accounts for both a great many failures and the occasionally dramatic story of great success from a humble start.
Although I wouldn't go so far as to think of it as an edge, I have learned to take my losses quickly (miss some great trades) and live to fight another day. An asset -- an inportant tool -- but not an edge. When you are light on capital you must master taking a quick, small loss and not dwelling on the ones that get away. The two realities go hand and hand.
All traders learn in one form or another the velocity of profits and losses as it affects their account balance. And too many traders discover their account balance is totally depleted even though they used âa recommended position sizeâ. This happens because these traders cannot put position sizing and account management into the same sentence. What traders often fail to understand when adopting position sizes is the frequency or rate of trading transactions must be a component of position sizing.
The â4% loss in a week being a fatal blowâ statement came from a conference I attended. It was a specific case. It does not mean to state categorically that if day traders have 4% losses in a week they are will be devastated.
The story goes some thing like this. When a group of us went out to dinner the trader who sat down next to me was grousing to me about his day trading of stocks. He was at the conference to learn some new techniques because in 12 weeks of day trading had lost about half of his account (averaged 4% losses a week) and was frustrated.
There were no talks at the conference I wanted to see the first thing the next morning so I offered to look over what he had done. When he showed me his trades it was apparent he was over trading, ignoring correct position sizes in relationship to his account, stopping his trading after strings of four losses in a row and not following his sketchy trading plan.
What that trader failed to appreciate was he was losing 16% of his account per month. He was totally ignoring the velocity of profits and losses. And in this case was over whelmed by losses. He did not understand the higher the rate of trading transactions applied to his account the smaller the position size he must using because the rate of losing.
He emailed me a few months later. He had cut his trading down to a quarter of the transactions he had previously taken. His position size was readjusted to each new weekly balance and losses were now in-line with the average number of losses in a week. His big complaint is that weeks that were profitable not big enough (because he had only half an account). But the plus side was the couple losing weeks he had had in a row were not the 8% loss he would have taken using the old methods.
Thank You
I never use back-testing to find cherry picked systems. I always start by finding a market inefficiency and then designing a system to fit it. I won't trade any system where I don't understand at a deep level why it works.
I find back testing to be of very little value for most of the systems I design. It's very hard to find a back-testing environment that will give you reasonable results for short term systems. For longer term stuff, backtesting is probably more useful.
Forward testing with a fake-money account has some value. It's good for debugging systems so that the trades you actually get are the ones you think you should have gotten looking at the chart afterwards. That sounds basic, but tends to be a problem whether a system is hand-traded or implemented via computer. Inevitably issues like thin markets, slippage etc. crop up that aren't obvious just looking at a chart.
Once the system is debugged, if the inefficiency is still there it's time to trade with real money. There's no substitute for the real thing, especially if your system requires you to move any kind of volume.
I find back testing to be of very little value for most of the systems I design. It's very hard to find a back-testing environment that will give you reasonable results for short term systems. For longer term stuff, backtesting is probably more useful.
Forward testing with a fake-money account has some value. It's good for debugging systems so that the trades you actually get are the ones you think you should have gotten looking at the chart afterwards. That sounds basic, but tends to be a problem whether a system is hand-traded or implemented via computer. Inevitably issues like thin markets, slippage etc. crop up that aren't obvious just looking at a chart.
Once the system is debugged, if the inefficiency is still there it's time to trade with real money. There's no substitute for the real thing, especially if your system requires you to move any kind of volume.
livevol_ophir
ET Sponsor
Quote from The Big D:
I never use back-testing to find cherry picked systems. I always start by finding a market inefficiency and then designing a system to fit it. I won't trade any system where I don't understand at a deep level why it works.
I find back testing to be of very little value for most of the systems I design. It's very hard to find a back-testing environment that will give you reasonable results for short term systems. For longer term stuff, backtesting is probably more useful.
Forward testing with a fake-money account has some value. It's good for debugging systems so that the trades you actually get are the ones you think you should have gotten looking at the chart afterwards. That sounds basic, but tends to be a problem whether a system is hand-traded or implemented via computer. Inevitably issues like thin markets, slippage etc. crop up that aren't obvious just looking at a chart.
Once the system is debugged, if the inefficiency is still there it's time to trade with real money. There's no substitute for the real thing, especially if your system requires you to move any kind of volume.
agreed...
Yes ... 4% per week is a disaster over a period of more than a couple of weeks. By week three common sense should have indicated he was trading to lose.
Quote from Rabbitone:
to learn some new techniques because in 12 weeks of day trading had lost about half of his account (averaged 4% losses a week) and was frustrated.
Thank You
8 pages later and not one person has approached the question you asked in the last paragraph...
The beauty of you being located in the UK or Europe is that you are able to access a spread setting firm/"bookie". That is not possible in the US. It provides you the ability to outlay a minimal amount to test your system live without breaking you.
Now, to the real world, trading ES...you should not even consider trading ES contracts until you have accumulated $10000 , the bare minimum to trade 1 contract.
I have seen your calls and trades and I sincerely hope that you have put a quid on your trades...you will be a winner if you continue the path you are following.
On other thing....can you refer to your performance to date and whatever backtesting/forward testing you had completed and see if you can also counter trade your system with the appropriate money management strategy your are using?
You do have a money management strategy , don't you....
:eek:
NiN
The beauty of you being located in the UK or Europe is that you are able to access a spread setting firm/"bookie". That is not possible in the US. It provides you the ability to outlay a minimal amount to test your system live without breaking you.
Now, to the real world, trading ES...you should not even consider trading ES contracts until you have accumulated $10000 , the bare minimum to trade 1 contract.
I have seen your calls and trades and I sincerely hope that you have put a quid on your trades...you will be a winner if you continue the path you are following.
On other thing....can you refer to your performance to date and whatever backtesting/forward testing you had completed and see if you can also counter trade your system with the appropriate money management strategy your are using?
You do have a money management strategy , don't you....
:eek:
NiN
Quote from livevol_ophir:
I work for them. You think it's odd that an employee believe in the product of their employer? I think it's odd that you think it's odd. Why would I work for them otherwise? Maybe I misunderstood?
It is odd that you would pay a sponsorship if you were an employee and not financially involved as an owner.