*Why bother? What’s the problem?*

It often seems that retail traders do not understand, let alone effectively apply, the concepts of risk/money management in their trading. By this I mean that they would struggle to answer the following questions:

- How much money should I risk on any one trade?
- How much money should I risk on trades that are open at the same time?
- What is the “statistical edge” (whether positive or negative) of my strategy?
- What is the volatility/variation of results for my strategy?
- How should the answers to (3) and (4) impact on my answer for (1)?
- What is the expectancy of my strategy?
- Why are transaction costs important?
- What are realistic return targets from my trading, over a month, over a year?

The reason why I am confident that most retail traders are unlikely to know that is because I am coming across more and more professional traders (working in banks and hedge funds) that don’t know the answers to these questions either.

I am not going to be providing my view on the answers to all these questions in this blog post. Instead I will provide some data on all the K5 January trades I took in January. This data should provide some food for thought for answering the questions above. Happy reading!

*Example of risk management in action*

Throughout January I have gradually increased the amount of money risked per trade, gradually making the way to £200/trade, as shown on the following graph:

This £200 is divided by the stop loss size (in pips/points) to determine the volume/quantity I should use for the trade. Thus, the stop loss size varies for each trade, the quantity traded varies for each trade, but the financial risk per trade is always about the same. For this past week the risk per trade has been about £175. Next week this will increase to around £200. How much should this figure be for any trader?

*So how much should you bet on each trade?*

Well, that depends on:

1. The size of your trading capital – do you have £10k, £500k or £5 million in your trading account/capital.

2. Your attitude in terms of risk-seeking/risk-avoidance, and general aggressiveness (which will in turn depend on whether you have another source of income, dependents, your age, etc)

3. The amount of edge of your trading style (for me the past month it has been 4.2%, see table below).

4. The volatility in the results of your trading style. Assuming that the edge across two strategies is the same, then if you have a low win rate but win really big when you win, then you should use less per trade. If you have a high win rate but less win per trade then you should bet more.

My suggestion would be to bet as small as possible until one is confident that they are trading with an edge. The purpose of that is to protect your trading capital until the point in time that you are consistently trading well, meaning that you have a positive expectancy every time you put on a trade. This is what I have been doing for two and a half years – my position size has varied between £50 to £250/trade, which is a very small portion (way less than 1%) of my available trading capital. Equally important, once you start trading well, you need to increase your position sizing – this is another common point that retail traders, as well as traders in hedge funds (or other trading entities) struggle with.

So, in a nutshell, a lot of this is dependent on the edge of a strategy as well as the volatility in the results.

*Hints for calculating “edge” – using my trading statistics for January 2016*

Here are two more sets of statistics that may help you think about these things – the first is an overview of how effective the 88 valid & executed setups have been for me in the month of January – did the valid setup correctly predict price movement (yes/no/sort of):

The second looks at the actual hard profit+loss for the month, broken down by financial instrument:

Explanation of the above chart: The second column (from the left) indicates the number of K5 trades taken in the instrument. The next column shows the total amount of R (where 1R = £200) risked for all those trades. Bearing in mind that the R risked per trade has gradually increased (see the post on risk/money management for more details on that). Then you have the amount of profit (or loss) expressed in R, and then the the amount of profit/loss in £ (which is the prior column multiplied by £200). Finally, and i guess the most important column is the return expressed as a proportion of the amount risked. I’d like to propose that this is an *approximation* of the strategy’s edge.

I did an equivalent calculation for the 68 trades I did in December (see bottom of that blog post).

Most people would laugh at an edge of 4.2% – that is most non-professional people. Any serious professional in the trading industry would sell their mother in exchange for a 4.2% edge. No, I don’t expect anyone reading this to agree with me on that point.

Pingback: 28/1 – Wrap-up for Week#4 | George's FX trading blog

nice to see you are trading 1R. Is the 4.2% edge inclusive of trading costs? i’m assuming not. What sort of % is your trading costs. I worked out my share trading costs are 1.5% annually, would be 2.2% inclusive of research data services – this to me is quite high, i would prefer something around 1% inclusive of research costs.

what do you consider a sufficient sample size to quantify an edge? 100? 500? 1,000?? What sort of sample size do your quants use?

i am also really surprised a bunch of quants don’t have statistical models for risk management, position size, trading costs and active specific risk in general. couldn’t you get some enormous retainer to do consulting for this type of thing??

LikeLike

Thanks for your comment James!

The 4.2% is after transaction costs associated with the trading (i.e. the bid-ask spread and commissions, if applicable). However it does not provide for any overhead costs such as office rent, computer equipment, subscriptions and so on. That said, those costs are (at present) not very high – and would be fairly minimal if I was trading with a higher value of R.

I think a sample of 500 trades, taken in a very consistent manner, would be reasonable. This implies it would take me 5-6 months to figure out whether I am really trading with a 4.2% edge or not. Though it would take less time/trades to figure whether I am trading profitably or unprofitably. 500 trades would not be an acceptable size for professional traders in hedge funds. They back-test their strategies over as much data as they can get their hands on.

In terms of quants having models etc – most of them do have models – however some individual traders within a fund, although having an understanding of the numbers and the maths, might struggle to apply it to their trading, i.e. putting on the size that they should be putting on based on their apparent edge – presumably psychology can often be to blame for that.

LikeLike

on that last paragraph – really surprised it isn’t automated to take the emotion out of it.

good luck with the talk this week – let me know how it goes – i’ll keep a closer eye on your blog from now on…. i’ve been catching up with your old posts and i definitely think you seemed more confident and have the belief you can be successful at trading.

LikeLiked by 1 person

Thanks James!

My talk for the hedge fund is still 2-3 weeks away but thanks for the good-luck wishes!

LikeLike

Pingback: 5/2 – Trading Wrap-up Week #5 | George's FX trading blog