Prediction Wizard

As our next  examination, we  would like to make some consideration about scaling  during potential  trending days, show a numeric example  and summarize our view on scaling.

Helping to demonstrate our view on this we created the following  drawing:

On this we have a trending chart,  and marked six very important  points on this chart from Point 1 – to  point 6.

The corresponding Price levels  of these six points are as follows:

1:  Ref + H1

2:  Ref + H2

3:  Ref + H3

..

6. Ref + H6

The horizontal axis is the time.

Assume, that we had a Bullish prediction for the day, the price pulled down to the Ref level,  that we considered a good Support level, and at that point we opened a Long position.

All of this happened during the first 5 – 15 minutes of the trading day.

The price starts to move up as we expected.

What to do next?  How to proceed with this  position. Add to that or close the  position  with a trailing stop or select  some other strategies.

The major difference between our playing strategy and the play of many other strategies of intuitive traders, that  we have a predictor, which gave us edge to open the long position early enough in the day.

At the time of the market open and after the first 10 – 15 minutes most of the cases we still do NOT have the high confidence decision  to make the distinction between a Trending day or a Range Day.

Though that would be crucial  information to make decision about the  existing long position.

As the time passes and we collect more and more real-time market information our confidence might increase, but  we might not reach  the necessary confidence level by the time, we must make a decision.

In some future  entries, I will  write a bit  about Range day identification and  trading day identification and about the probabilities to be in either of them.

As we mentioned in the previous entries in these series we need to have high probability price target at  far enough levels from the current price  to  add to existing positions to be consistently successful  in the game.

At  point “1”  we still don’t know that we are in a trending day or a range day, even if our  estimates is like above 60% leaning to one way.

At point “2”  or soon after that  we can confirm, that we experience a pullback in an Up-trend, or a Reversal in a Range day.  If more than 68% of H1 up-move retraced, than we experience  a relatively high probability range day.

The point, that we are experiencing a trending day  could be confirmed with the highest probability at or above point3.

Though it is possible to make a good guess at points 2 about the distinction of pullback / breakdown  but only very experienced traders should make that call, who also extremely quick at decision making.

We also know the trending characteristics, and  the attributes of a strong trend.

During extremely strong trends we experience very shallow pullbacks and see few  red bars / candles on the chart. (As if someone would pull a rope from the top of the curve with high force.)

We also know that the higher the level we add to the existing trade the less probability we will win with that position.

To  solve these  seemingly contradictory requirements (To reach higher profit levels in a trending day and not to lose the existing profit level at point1 in case of a range market, when we reach point “1”  in the first 10 – 20 minutes of the trading day.)  we carefully analyze the current  market both at our trading time frame and one level below  that time-frame.

What we plan to execute is  to sell at the top of the first move at the momentum high, and  waiting for a pullback, BUT immediately  put in a Buy Stop order to  buy it back, once the price moves  above the  level, defined in point “3”, the previous intraday local high.

But we do not do that without some other consideration, based on market action.

Obviously we have price target levels  available before the market opens, and we know the potential resistance levels  above  current market level.

If  at point “1”  we already  reached  more than 40% of the  daily price target (Which could be calculated using the previous days, and  fine tuned by volatility and other market behavior.)

then  the  Long position at point “3” will be smaller, but could be still bigger than the first long position, opened  at Ref level. If already reached higher level (50 – 60% of the daily price target) than the position size at Point “3” could  be even smaller.

Within a strong  trending day we might have as much as 5 – 6 up-legs or even more with pullbacks we do not open new position above point “3”,   only try to manage the closure of that position. Usually close in one or two installment.

When we considered this execution strategy  we did not want to make  it too complicated and also agree with Linda Raschke  to open  one bigger position at low enough level.

When we reach point “4”,  we already know that  trending is in effect, and  use a bit different exit rules to close existing long position.

The beauty of the market that we do not have identical days. Every  time, every trade a little  bit different, but if we look close enough every day assembled  by the same, limited number of puzzles (patterns) that we need to discover  and exploit  during our trading career.

To highlight  the importance of scaling and position size we present the following numerical example.

Assume we have two traders, Player A and Player B

Both of our traders use  exactly the

-          Same asset class for trading

-          The Same time-frame

-          Both of them uses the predictor

-          Both of them uses the same implemented strategy on top of the predictor, which leads to the same  time of entry / exit.

The only difference between the two sample players is  the position size, they select to play the signals.

Assume that the played strategy   on the selected settings leads to  a  55% Win ratio,  and the size of the winners are equal to the size of the losers. (This might happen, if we  hold positions overnight.)

Both of our traders playing with a 100K account  (Can use  margin though)  and  both of them makes 100 trades.

Player  A: selects  the position size in a way, that he  is risking 1000$, and also accepts 1000$ as profit.

With these the results of Player A after 100 trades, without slippage and commission:

Win:  55 tradeas, 55K

Loss: 45 trades, 45 K

Win – Loss = 10K

The second player is an idealized player, who  is playing the game with dynamic position size adjustment.

He selects the high probability trades to be 50% bigger size (150K, with margin) , and selects the low probability trades to be 50% of the original size, playing with 50K

Assume the ideal situation, that all  high probability trades, played with increased size will be  the winners, and all the low probability trades, playing with lover size will be the losers.

With these:

Win:  1.5 *  55 = 82.5 K

Loss: 0.5 * 45 =  22.5K

Win – Loss = 60 K

Obviously not all high probability trades will be winners and not all low probability trades will be losers.

But considering that the two players played almost exactly the same game,  the difference is Huge.

We would call this the 60 / 10 Rule.

The longer term results of the player, who successfully adjust the position size  according to win probability / conviction in the trade,  might reach much higher profit levels.

Even if we manage to realize  much less than the actual difference above, say instead of 10K we manage to reach 15K with those 100 trades, it would still be big difference compared to  the strategy, when we use a fixed position size for all of our trades.

Very  often  what differentiating the losers and the winners is the dynamic position size adjustment.

But this is  definitely not for beginners. Without  a true edge, we can’t  even start talking about  position management, so we must establish that first.

Obviously we can set the difference between the  different position size more flexibly or less aggressively.

In our example we had  a position size difference of  three to one  between the high  probability trade size and the low probability trade size.

Predictionwizard try to give us some support  with the presented win probabilities.

Finally to summarize our view on position management:

1.       Position management and scaling is  a very effective method to increase the profitability of the professional trader  who knows when to push the gas pedal, and knows when to scale back or even not  participate  in a trade.

2.       Position management and scaling  could be extremely dangerous and potentially a way to destroy the wealth of the less experienced trader, who  might push the gas pedal at the wrong time,  and also scale back at the wrong time.

       In case of the above example if a trader opened  all low probability trades

       with a position size that is three times the size of the high probability

       trades, than:

Win:  0.5 *  55 = 27.5 K

Loss: 1.5 * 45 =  67.5K

Win – Loss = - 40 K

        Total loss:  40 K

        So this player have a real edge in terms of win probability, but execute

        poorly with wrong  position management and as a consequence

        losing his short in the game.

3.       Scaling and position management is strongly related to  knowing the win probabilities available in a given situation.

Most of the intuitive traders  do not know their win probabilities, and their profit ratio that leads to less confidence in the heat of the battle and that leads to outsized losses.

       When position management / scaling  properly implemented, it  could be

       the crown jewel in the market participation,  but strictly suggested for the

       very experienced  traders, who are already consistent winners in this game.