If you went short in March 2009, would you still be short with no stop? If so, you'd have about a 35% drawdown on that 1 contract per 100K and no way to know when to ever close the trade. Unless you would just let the contract expire at rollover or something, in which case you've used a time stop. No stop only works when the market you are trading is guaranteed not to embark on a multi-year move against your position. Since no market meets that criterion, you have to use stops. What compels you to give this terrible advice? It's one thing to make provocative statements, but it's another thing to make provocative statements with no thought behind them.
Well it depends. If he gradually increases the position, having appropriately calibrated respect to capital, he will almost certainly find a reversal big enough to take him out of the trade ... mkt does not go from point to point in a straight line ... especially ETFs and FUTs
While that is a fair point, that's not what he said. But, if he does that, he's basically Martingaling, which is a terrible money management strategy in almost every context. In fact, the one context where Martingaling can make sense is when you have a hard stop but the market you are trading is not liquid enough to handle your entire position at one go, so you can average in at better prices if you take heat on the trade. If you are right and the market goes your way, you get a better average price, although you'd probably get some slippage on the exit because your position size would be some multiple of what the market liquidity can bear, so you'd have to factor that into your decision to Martingale or not. Still, you'd exit all at that hard stop. But Martingaling without a stop makes no sense to me. Think about it, you've already been wrong once with your initial entry, what makes you think you'll be right when you add? A lot of what seems to be the impetus behind not using stops is a desire to never be wrong, which is completely impossible in trading.
Well "martingaling" implies an exponential growth. One can instead increase linearly or logarithmically and most of the time pull it put. Especially if this is done within a large folio, with good hedging rules (for instance blocking through cointegrations, to focus the investment) and small starting sizes. The massive scalping across the folio action would normally self-sustain the bigger drawdowns (which somehow would be the "real" trades, in the sense of done with a "good" size, at a good cost basis).
Yeah, that works sometimes, but you are not always going to be able to average in to get exactly 3 points away from the market price. And, what is this about taking a loss? Isn't that using a stop? See, your advice only works if the market stays within a range. Since that is not always going to be the situation, it's bad advice. You never EVER want to put yourself in a situation where the market's "fat tails" are virtually guaranteed to work against you at some point, unless you've worked out the long-term probabilities to such an extent that you know you can get hit with one and still come out ahead in the long run. But, given that the market since March 2009 has basically been a 700 point fat tail relative to the "no stops" strategy, it would take you 700 1-point wins to make up for that one loss. Is that likely to happen? Not really. There are obviously things beyond a trader's control where a "fat tail" can come along out of the blue, but setting up a rule that eventually puts you in the crosshairs of a "fat tail" is just stupid.
OK, right, not necessarily exponential growth, but even linear requires that you get a 50% retracement from your initial entry to your second entry and in a strong (or weak) market, 50% retracements are not guaranteed. I'd rather just set an initial stop that rarely gets hit and then move that stop as the market action progresses. Maybe because my initial stop gets hit so infrequently, it doesn't bother me when it does. I recently went a year without my initial stop getting hit on ES trades and then it happened at the end of May. Of course, I worried that it might mean something was wrong, but it hasn't happened since then and June was my best month of the year so far. Where I put all my effort is in defining which trade set-ups might result in an initial stop getting hit. I've identified the most relevant parameters and can filter out the ones that will get stopped out that way. All it took was one parameter and I identified trade set-ups which are 37X more likely to have the initial stop get hit (37% of the time for that specific set-up vs. 1% for set-ups with a different value for that parameter). Once I started avoiding those trades, there were no problems. Seems easier than trying to add, scalp and hope.
What if the trader is no better at "massive scalping" than he is at cannonballing a losing position marking wildly against him ? What turns exponential is the capital requirements - and the trader's stomach lining erosion rates.
Problem with stops, is that their *systematic* use (not occasional, discretionary) leads to strategies which, at best, fluctuate around 0 PNL. That is, they are just "entertainment", until a negative drift or some error takes you away. The key is in the overall architecture, capital, large automation, and smart strategy that guarantees good hedging and controlled investment... it's nothing really easy to accomplish for the average joe ... it takes many years of hard work even for top minds and specialized knowledge ...