The scaling in/out section in babypips school misses the mark

Babypips,

With respect, your “coolest ever guide to scaling in and out of your trades” in the babypips school misses the mark.

Your summary states that “scaling can help you adjust your overall risk, lock in profits, or maximize your profit potential”. However, none of this offers any mathematical benefit. I’ll attempt to explain why.

The illusion with scaling occurs — and indeed many other forms of MM — because people tend to think of all of the component positions as somehow comprising a single trade. But the fact is that each component either adds to, or subtracts from, eventual overall bottom line, and must be on-balance profitable in its own right in order to be justifiable. Grouping individual components together is nothing more than an arbitrary process.

For example, one school of thought is that, if I have locked in $100 on a trade thus far, then I can afford to risk a further $100 as I’m guaranteed not to lose overall. But the reality is that the second component entails a risk all of its own, in that it will either increase or decrease my total account balance, regardless of whether I’m +100 or –100 when I enter the position. It carries exactly the same risk even if I hadn’t entered the first position at all.

Scaling is nothing more than a type of averaging, or ‘hedging one’s bets’. As such, it can smooth out the bumps in a trader’s equity curve — just as any kind of more frequent trading does — but it can not, in itself, improve a trader’s overall expectancy.

Scaling out provides exactly the same illusion, and smoothing. In a strongly trending market, a trader will, on average, optimize his expectancy by letting the whole position run as long as possible. In a sideways market, or if one is trading counter-trend, then taking profit early on the whole position is preferable. The strategy should be determined by market probabilities and behavior, not the trader’s P/L hopes and expectations, which the market takes no cognizance of. If the trader is uncertain of market probabilities, then it matters little whether he exits earlier or later, because 50% of the time one approach will outperform the other, and vice versa. Hence over a large enough number of trades, the end result will be the same. Without an edge, the trader is merely guessing, and therefore gambling.

Scaling is effectively just another form of MM, and all MM ultimately does is re-balance return and risk, and/or redistribute wins and losses. The only way MM could improve overall expectancy is if the trader knows in advance which trades have a higher probability of success, and then staking more on those trades. That is true for any kind of gaming, or indeed any activity that involves mathematical uncertainty.

Much the same can be said for other MM techniques like, for example, advancing the SL to breakeven. Unless there is technical or statistical validity for doing so, it represents nothing more than a ‘feel good’ measure. The real question is ultimately whether, on average over a great number of trades, the reduction in risk compensates for the number of would-have-been-profitable trades that are subsequently cut off prematurely because they were allowed less room to move. Again, any edge comes from understanding and exploiting market probabilities and behavior, as opposed to focusing on ‘locking in profit’, or some other kind of P/L-based criteria.

A similar illusion that leads to wrong thinking is the widely held notion that floating P/L and realized P/L are somehow different. It is easy to show that they are the same. If I close my current position and then immediately open an identical one, I have realized my P/L on the closed trade, but my account balance (if we disregard the additional spread cost), exposure to risk, and available margin all remain exactly the same. And the notion that realizing a loss removes the opportunity for the position to return to profit is equally flawed, for two reasons: (i) one could simply open another position; and (ii) unless the trader has some kind of edge, the floating loss has a 50/50 probability of improving, or worsening, henceforth.

Let’s suppose that I’m currently long and I happen to somehow know that there’s a > 50% probability that price will continue to rise. Should I keep my position open? Of course I should. But if I know that the probability is less than 50%, then I should close the position (and in fact, go short). The P/L in the trade to date is irrelevant. What has already happened can’t be changed; at every point along the way, I can improve or worsen my bottom line by the decision I make now, and whether price rises or falls henceforth.

This leads us to the (correct, I believe) conclusion that risk is a dynamic process. When I originally open a buy trade, my TP might be 60 pips away from entry, and my SL 20 pips. Hence, if I perform no trade management henceforth, the trade supposedly offers 3:1 return-to-risk. Now let’s say that price rises 50 pips. Now price is 10 pips away from the TP, and 70 pips away from the SL; hence the risk at this point is now 7 times the possible return. That would be precisely the situation if I was to close my position and immediately re-open it. Hence I would argue that risk/return analysis is much less valuable than the textbooks would have us believe.

If (and I stress, IF) price movement was a completely random walk, then P/L would be nothing more than a fortuitous exercise. Win rate, and average win-to-loss size (R:R), would be exactly inversely proportional to each other. In other words, everything else being equal, doubling our return-to-risk on each trade (for example) would halve our win rate, leaving overall expectancy unchanged. The key, then is to find inefficiencies, or non-randomness, in price movement, that somehow overcome this inherent equilibrium. In other words, an edge comes from one’s ability to find situations that offer either directional or behavioral bias.

Conclusion: There is only one way to be profitable in forex, and that is to be net long while price is rising, and net short while it’s falling, heavily/frequently enough to overcome costs (spread, swap, commissions). The more accurately we adjust our net position relative to turning points in the market, the more pips we bank. MM techniques like hedging, scaling, varying position sizes, etc merely re-balance risk and return; they can smooth a trader’s income, but they can never, in themselves, improve expectancy, and hence long-term P/L.

The attached XLS was created originally to highlight the folly of Martingale, but it can equally be used to demonstrate, that no matter what MM is used, total expectancy remains unchanged; and hence a winning methodology comes from one’s ‘gameplay’ — in forex that equates to the efficacy in timing one’s entries and exits.

MM spreadsheet.zip (31.3 KB)

I see it just took you six years to make your first post here… but let me say it was worth the wait.

You share a user name with a person over at ForexFactory - any relation?

Many thanks for the kind words. :slight_smile:

Yes, I am also hanover at Forex Factory (and a few other forums also).