You thought that the prior months close was germane to the trading signal.
Yea, in that example I was still stuck on the one month lookback period discussed above, which I realize is not what you are using. Fortunately the math works out about the same in this case, given flat recent returns, as you yourself pointed out.
That doesn't address the criticism I was trying to make, though, that your strategy would perform poorly in a market that dips suddenly, triggering a trade, then recovers suddenly. I think that's a very possible outcome in this case.
Your counter example of it performing well if this is the beginning of a long bear market only illustrates this point. The DM strategy is counting on the signal being indicative of a long bear, so it does well if it is and poorly of it isn't.
You blamed a strategy that trades less than 2X per year and at most once per month on average and by a small minority of individual investors with rapid intraday swings of asset prices.
I think this is a fair defense in practice, if maybe not in a theoretical sense. Your argument that other actors are doing more harm than you are does not absolve you of the harm you do, does it? If HFTs are Pol Pot, you're just like a run of the mill occasional hit man?
I've yet to see any explanation for why momentum traders don't amplify volatility by piling on to current price trends. You can think your effect is minimal, but I think you have to agree the effect exists.
Second. Why would I worry about what the closing price of assets was at the end of the month, when knowing that now would have absolutely no effect on my behavior?
This is probably just personal preference, but I like to know what my future options are going to be. My savings rate isn't going to change today based in my current expectations for market returns, but I still have a forecasting spreadsheet and I'm not even making trading decisions. You know you'll be making trading decisions, so I don't understand the mindset of not trying to forecast what they will be.
you only ask about the future performance of DM on days where the market moves against it. If you made speculative comments on days like today
Wait, are you accusing me of not discussing this topic today? Because I'm pretty sure it's today right now, and here I am.
DM strategy performance had little to do with today's moves, as you know, and everything to do with the motion relative to the price at which you sold it. You sold out on Sept 1 at 1913.85. Today's 1.6% drop gets it back down to 1958.08, so you only need to see another 2.6% drop in the index before you buy back in to come out ahead. If it doesn't drop at least 2.6% from today's price before you buy in, then this will have been a bad time for the DM strategy. You won't get a buy in signal until the index at least crosses the price at your six month lookback period, which was never below 2059 for the past six months, so you won't buy back into stocks for at least six more months unless the index exceeds that price. See why I want to forecast?
If that does happen in the next six months, you will have locked in losses of at minimum 7.5%. More likely, I think, is that your signal won't trigger a trade until six months out from the Sept 1 price of 1914, since I think it more likely it will be above 1914 on March 1 2016 than above 2059 on October 1, or 2108 on November 1, or 2111 on December 1, etc. (Prices determined by six month lookback).
The flip side, and what you're hoping for, is that the market crashes from here by more than it crashed before you got the signal to sell (about 10%). If that happens, you will have successfully avoided losses and you hope you get the signal to buy back in again before the next bull gets started by rising more than the losses you avoided. That's always the trade off here, trying to avoid losses but not avoid the gains that follow.
Am I misunderstanding any of that? You keep accusing me of misunderstanding the details, so I'm trying to lay it all out there for everyone to see.