Monday 17 January 2011

Trading Every Day to Make a Living: Part 3

The way I set up my auto trader, Flo, for Hybrid Trading is very different to the parameters for my algos of the completely unattended auto trading. While the methodology is the same, the way that it is managed is different. There, I'm after the best return on investment, unattended.

There was a comment to Friday's part 2 post saying: "Your losers are three times as big as your winners so wouldn't the drawdowns (troughs) be large?" The annual maximum drawdown ran at less than $2,000 per contract traded as a fully automated basis and on a managed basis I estimated it to be half that. By using HT, the average losses are NOT three times as big as my winners. That's the point of HT. But even if they were, the high win rate makes the overall method profitable, as long as the winning trades are traded to a reasonable percentage of their potential.

For HT, I am after a high win rate only, with optimised drop dead stops and optimised first scale out points.  The aim of HT is to increase profitability from that base using discretionary but rule based trade management.

I put the above in bold because it's important to understand this. Every trade's outcome is different, although the general entry criteria may be the same. It's what happens after the entry that determines the outcome. I look for trade entries that fit my Trading Plan. However, I cannot program in all the context that, as a discretionary trader, I use. So Flo enters in HT mode with an optimised drop dead stop and optimised or money related first scale out target. I then have to manage the trade.

Some trades I scratch straight away. For example, a long just below a strong resistance and that has the odds against the trade, I scratch. I manage my losing trades so as not to lose the whole drop dead stop. I typically have a drop dead stop of around  $300/contract for the ES in case I need to double down. But I rarely use it. The market tells me to get out much earlier, but as it's different for every trade and not a fixed number, I can't as yet program all the possibilities. Also, using HT, I can add trades that Flo misses.

The same goes for profit targets. Every trade is different. Using a fixed profit target would limit my profitability, so I have a first scale programmed where I exit at least a third, but I manage the rest. I have a version of Flo HT with scaling out built in that I use when I'm not paying attention. It does well but not as well as when I manage manually. Again, because every trade is different, I manage my exits to make sure that I can earn the most after assuming the risk of the trade.

When setting up Flo to run in a purely automated way, the result is that the win rate is lower and the trade profile is very different. 

HT can be used in many ways but I see a lot of merit in it for many traders, as HT helps get into the trades in a consistent and disciplined manner. This is a big component of what is required for Hybrid Trading. Managing the trades after entry can be done using strict rules that are consistently applied.

Beginning traders can start with a more robotic approach, having programmed stops and scaled out exits, and as their skill level improves with screen time, they can slowly take more hands on control.


On Saturday, I sent the screen shot above to one of my students. It is Friday's ES RTH trading. This is a good example of HT . As I said, We have programmed Flo to achieve fairly high win rates and profitability, yet there are times when I override her by scratching a trade, moving profit targets up or down, or moving stops closer (never further).

The question I have been asked is "Why?" If Flo is doing OK, why do I mess with her? There is an opportunity cost to trading. If I'm risking my money, I want to maximise the return.

Looking at Friday's ES, the key to the RTH session was the Gap trade. I tweeted it a few times on previous days to draw your attention to this great type of opportunity. Friday's Gap was especially low risk and high reward for me because of the Fib point below that fit the MP context so perfectly. It was really easy to see if I had been wrong.

What followed from that successful Gap/Fib entry was that I had great confidence in Flo's long trades after that. In fact I increased size and did not exit the extra lots, as per Flo's all out exit, but rather scaled as I do manually and then reloaded at the next Flo long signal. The end result was a pretty awesome green day.

Looking at Flo's results, they were great. 100% for 2 trades, 2 ES points. Again, win rate is king. With a high win rate I can trade size with confidence, as the drawdown can be managed. With a low win rate the drawdowns can bury you without a trace and your account can be empty by the time the winning trades decide to appear and match your stats. The basis of what I teach students to get to CP is the maintenance of this high win rate. The psychological and financial criticality of this aspect of my strategy should not be under rated.

For me, HT is definitely the best of both worlds.

Holiday in the U.S. I traded a bit of Euro futures in the London morning but taking the rest of the day off.

2 comments:

  1. Hello,

    In Part #1 you mentioned that you would show us: "How to lock in that $250 per day."

    Is it in this section and I don't understand it? Or does that come in a later post?

    Thanks for sharing your experience with us.

    ReplyDelete
  2. Anon 18:51, still writing that part, Part 4, later this week.

    ReplyDelete