using only historic price data to trade

[davao_danny]

Hi everyone.

I'm completely new to trading, and plan to use very simple Excel-based running charts of historic prices as indicators for buying long, selling short and closing positions for individual companies (starting with the FTSE 100, then expanding to other markets).

I intend to make trades of around £1000. Assuming a (possibly pessimistic) cost of £20 per trade, I'd need to make an average gain of more than 4% for each position to produce a profit.

1) Has anyone successfully used only historic price data to trade (ignoring all other factors such as profit forecasts and politics)?

2) Can anyone suggest how many running days of data I should use to find trends (e.g. the last 90, 180 or 360 days)?

3) I'd also like to hear people's views on whether to go with consitent trends (i.e. a reasonably linear graph) or to buy/sell based on more extreme sudden changes in prices.

4) Which online trading services are reliable, simple to use, and offer good customer-support? I will not be trading on margins.

Thaks in advance for any feedback.
Danny

 

[Andy_S]

Hi there

Just a couple of questions

1. Do you have a specific profit target in mind
2. What sort of timescale are we talking.
3. How many trades are you planning per day/week

If you are looking to make quick money then short term trend favours. If you can afford to hold the position for a higher profit then long term.

Just my humble opinion, I am sure there are others with more expertise in this area.

 

[davao_danny]

Hi Andy. Thanks for your input.

I've got no target, other than an average gain of at least 4%, to avoid losses.
Depending on trends, I guess most positions would be held for between a few weeks and a few months (and would imediately close any position that started to go the wrong way, rather than hoping it recovers).
Again, I can only guess that I'd be making just a few trades per week.

 

[plamar]

as for online trading services you may try FXDD. They are pretty easy to use and also have good customers support

 

[davao_danny]

Thanks for your advice, Plamar.

 

[convertyourbond]

You can create a trading system with just using historical data but a couple of things to consider:

1) key reports, earnings announcements, NFP's etc - keep your eye out on those and if you are using a HP only strategy then you probably shouldnt trade right before/after these.

2) you have to have a better plan for take profits and stop losses than just I want to make 4% on each trade.

The key thing about creating a mechanical trading plan is to FULLY believe that your system has an edge over the long run, so that when you lose money on a trade it doesnt affect you or rock your confidence in the system. That is the problem i had when i tried HP systems, my mindwould just get too skeptical about the idea that HP do not represent the future.

 

[tomorton]

1) Has anyone successfully used only historic price data to trade (ignoring all other factors such as profit forecasts and politics)?
If you mean using price charts, yes, most people on this site are doing this. Most fundamental data such as financial reports, broker forecasts, press tips and politics should be firmly ignored when trading this way: but as has been said, if a significant announcement is due, better to be in cash.

2) Can anyone suggest how many running days of data I should use to find trends (e.g. the last 90, 180 or 360 days)?
There is no single answer to this but a general guideline would be use just enough data to confirm that a trend has occurred, and its length should be relative to your proposed holding period - if you're likely to hold a share for only 3 days, then its performance over the last 360 is probably not a critical factor.

3) I'd also like to hear people's views on whether to go with consitent trends (i.e. a reasonably linear graph) or to buy/sell based on more extreme sudden changes in prices.
Both are reasonable choices. A linear trend is likely to continue tomorrow but may not give much reward: a more volatile share may move further in the same time but you run the risk of a sudden reversal.

Overall, when looking at a possible position you should -
calculate the cost to find out the decision to get in was wrong and you need to get out
not risk too much capital per trade - so that if you have a string of say 10 wrong decisions, you will still not get wiped out
be able to see where you will get out if the price goes up or down: charts should really help here, if you can't see both exits, don't go in
always know your stop-loss and always obey it
focus on how much you could lose, not how much you could make.

Good luck

Tom

 

[davao_danny]

Thanks very much for your advice, convertyourbond / Tom

 

[fayalac]

Hi danny.
First of all good luck, you are making a great step, but....., think twice.
I would start with demo accounts, thats no real money and if you lose you can realize and study the reasons of those losses.
You cannot start a trading day thinking on the profit you need, take from the markets what the markets are able to give, if less.... less, if more..., more, if your target is 4% from each trade you can be risking yourself to a tragedy in your account.
By the other hand historical data is important but what really matters is were the markets are heading in these moment despite to where it should be according to historical data.
Pay attention to what tomorton said.
Look back if you want but keep your eye ahead.
Good luck

 

[davao_danny]

More great advice. Thanks for your help, Fayalac.