BlueHorseshoe

MAE and MFE stop and targets are well worth adding to your list, I think. Provided market conditions don't differ wildly from the test data then they often hold up suprisingly well. A cruder version of this is simply an optimised dollar stop and target, though all the usual caveats about over-optimisation apply, and you want to see acceptable results across a broad range of parameters, not just one dollar stop/target that happens to work well when neighbouring parameters don't. A fast moving average can work very well for targets for mean-reversion style trading, but this leaves unanswered the question of where to place a stop. Actually, although price always reverts to some mean, but we can never know which one, then multiple MA targets may be an interesting idea to explore for scaling out. I might try and investigate this and report back . . .

Hi Roztom, I'm probably not doing a very good job of explaining myself. When I develop a strategy I do assume the worst case scenario. So my question about fills is for 'information' only - I just like to be aware the range of possibilites surrounding what I'm doing. Bluehorseshoe

This is just a guess, but going on the fact that there is a difference between the two platforms, I would guess that this may be to do with how you have the TS platform set up. There are options within tradestation whereby the limit order can be held on the exchange's order books, or where it can be held on tradestation's server until the limit price is hit, or where it can be held on your own computer until the limit price is hit. I'm guessing that you are set up with one of the latter two options and not the first. This means that your order doesn't reach the exchange until other limits at your desired price have begun to fill, and you get thrown in at the back of the queue. I hope that's correct/helpful.

Simply selling at H[1] and buying at L[1] on a stop and reverse principle would be immensly profitable if only it were possible to guarantee a fill at limit in every instance. Even though this obviously isn't workable as a trading strategy due to the fact that these limits are unfilled often enough to produce a negative performance, I still regard this as strong evidence for the inherently mean reverting nature of the ES.

Have you looked at more complex money management approaches such as Optimal F? Bluehorseshoe.

Hi SIUYA, Thanks for your helpful (and patient) reply. I see what you're saying above - but surely there might be possibilities to hedge in other derivatives? Something like the ES remains reasonably liquid even through a 'black swan' event, so you'd get a sell stop executed, even though perhaps with considerable slippage that's better than remaining un-hedged. I fully appreciate that my suggestion above is obviously nonsense - I'm just asking a stupid question to try and elicit an answer, not because I'm arguing the point in any serious way. Cheers, Bluehorseshoe

Hi TeamTrader, I could be mistaken, but I think guaranteed stops are just a feature of spreadbetting for us here in the UK - they're certainly not an option for exchange traded derivatives like futures. In fact, this is the reason why a spreadbetting company need to charge a premium; they're hedging themselves in an underlying market (usually futures), and there is no way for them to obtain a guaranteed stop therein. Bluehorseshoe

Hi Roztom, Thanks for your reply. My understanding is exactly as you describe, and when I becktest I always do require price to trade through a limit order before considering it as filled. So I test with the assumption of worst conditions, as you suggest. However, my purpose with this thread is to enquire how much worse might the backtest conditions be than what would actually have been realised trading the same (historical) market data. Knowing the worst case scenario is obviously the essential piece of information for system design, but having a way to estimate a best case scenario with limit orders would also be useful. I then know a range of likely outcomes for trading in real time. There are certainly circumstances in which a fill would be very likely at limit. For instance, suppose that I place a sell limit order today at a level three hundred points above the close. Now suppose that price does not trade up to that level until two years hence. Chances are (I imagine) that I'm front of the queue, even though thousands of orders may be queued there by the time the market trades to that level. There simply won't be many limit orders sitting on the order books for two years at any particular price level. Many thanks if anyone is able to provide any more definite info on this . . . Bluehorseshoe

Which is why true 'trend following' systems tend to be stop and reverse, and always in the market, rather than the more complex breakout strategies associated with the Turtles etc. I would suggest, however, that compounding of returns is also a key part of the 'edge' that trend-followers exploit - if you take away this aspect their returns are invariably dismal.

I don't know anything about volume profile - can you point me to any good threads to learn more about how it's used? Is it as 'complicated' as MP? Thanks.

Does anybody know of any study about just how far in advance one needs to place a limit order to be pretty much ensured of a position very near to the front of the order queue? When I backtest intraday strategies, with orders entered sometimes just minutes before a potential fill, then I require price to trade through my limit order before the order is considered to have been executed by the backtest. I am backtesting a swing trade strategy where orders are placed a day before, and I'm wondering whether I need to do the same, or whether placing an order this early makes a fill at limit very highly likely? Oh, and I'm looking at the @ES contract. Many thanks. Bluehorseshoe.

This sounds awesome! Thanks for sharing it with us. Isn't it a great coincidence that just a short while after posting your question you found the perfect answer to it . . and you were willing to share it with us all . . . including the retail price . . . and the free marketing promotion . . . Great stuff! I'm so pleased that this company is willing to share their fantastic cybertrader with the likes of us that I'm going to head over there and offer to pay double right now . . .

I don't really understand options well enough to know why those who write them can't cover themselves when the market begins to move against them - obviously it's not straighforward or everyone would be doing it. Part of your experience was obviously unlucky timing - I think many people go years doing what you were doing before it catches up with them, and in that time they pressumably have a great lifestyle. Just out of curiosity, what happens if you default on an uncovered option - or do margin requirements make that impossible? Anyway, I think it's great that you're still around and trading ten years later. Bluehorseshoe

Suitability to psychological makeup is a good question. I think it's often more of a compromise than many traders make out (like in the market wizards interviews where they always have a style that is perfectly suited to their personality - really?). For example, I hate losing trades. This is not because I have a social hang-up about being a 'winner' rather than a 'loser'. It's because I know that every single system failure starts with a loss, and system failure is what I live in dread of. One obvious way to counter this would be to have a much more short term and higher frequency trading style. Then my periods between winning trades might be just a few hours, rather than weeks. At the same time, I know from experience that I am not well suited to daytrading for a whole host of reasons (tampering with positions, screen boredom, the need to work a full time job etc). So really what I am doing is a compromise - its the 'best fit' that I've found for my personality so far, but I wouldn't say it's perfectly matched. As for the trading method you mentioned, then I see no reason why it shouldn't work. And I would have thought that the ES would be an instrument where it ought to be most likely to work well. There is a recent JSwanson thread about buying weakness and selling strength in the ES which might be of interest to you, as this is essentially what you're doing when you scale into a long at an excursion to a lower band. Out of interest, what bands are you using?

I trade without a stop or target as well (though I don't scalp, in any sense of the word), so I agree there. I understand the point you're making about probability and expectancy, and I know just enough about options to understand what you're saying. However, I think that people on this thread are often talking at cross-purposes (I posted to this effect early on) - they're using the term 'scalping' both to refer to trading for the spread (like a true, traditional scalper might, requiring low fees etc), and also for day-trading from short timeframe charts. With the latter, there are certainly many strategies where having a larger stop than target makes sense, as this is simply the only way to get a stop inside intraday volatility and keep a stop outside of it. I wouldn't tend to call these people scalpers myself though. While I agree that most strategies with larger stops will eventually experience some sort of 'black swan' event, I think most fail for a whole bunch of other reasons before they even get to that stage. Incidentally, if it's not too difficult to talk about, I'd be really interested to hear about your experiences with writing options.

Now why didn't I think of that? Thanks!

Hi Onesmith, The strategy is for trading the ES contract, though I generally test such things against other closely correlated instruments (ES.D, Spiders, YM, Big S&P, Diamonds, NQ etc) as a measure of robustness. In other words, I would expect it to have a positive expectancy regardless of whether the Globex session were included in the testing. It aims to take advantage of the 'window dressing' that is common around the end/start of the month, and therefore provides for long entries only. The entry is made intraday once other conditions have been met (in addition to the end of month condition that I am struggling to code).

You seem to contradict yourself here: having a target less than a stop is much closer to having a target and no stop than it is to having a target and a stop that is less than it. If you keep making a stop-loss bigger, in the end it becomes so big that to all intents and purposes it isn't there. Does that make sense?

I would recommend that you research various trading methodoligies and find the style that suits you best. Assuming you then want to trade a single market, find the market for which this type of approach is most suitable. For instance, I would be asking myself, do I want to trade breakouts (perhaps the Yen would be good), intraday trends (maybe the British Pound is your market), mean reversion (works well in many places, but none beats the S&P), momentum (perhaps the Euro), buy pullbacks (what about bonds)? Many people seem to become attached to a particular market for no sensible reason. Or because they like the glamour of it (some Forex and Gold traders). I trade the S&Ps. I hate the S&Ps - five hundred faceless stocks in some some country on the other side of the world - but this is the market that facilitates my outlook as a trader, and in which (historically) I would have been most profitable. In other words, you would do well to keep ego out of the equation when you choose a market and think only about your potential to profit in it. As for whther to trade multiple markets, that's down to you. I know that I would struggle to do this,as I tend to suffer from 'tunnel vision' in whatever I do, and can't juggle multiple processes well in my head. But many traders thrive on this challenge, and I suspect that most off-the-floor institutional traders probably trade multiple markets. Hope that's useful to you.

Hi Roztom, I 'swing trade', using daily bars and end-of-day order entry. However, if you were to look at my average winning trade, it would look more like what you might expect from daytrading. I never swing for the fences or ride trends like many swing traders do, and my typical holding time is only a few days. I'm currently researching a strategy that incorporates a fixed dollar stop and target (not my usual style). Once again, this uses daily bars and EOD entries for the ES, but the target is only 6 points and the stop loss 10 points - so it starts to look much more like a daytrading strategy than anything else . . I'm not sure what point I'm making, other than that things are never clear cut, and phrases like 'swing trading' and 'day trading' are simply convenient umbrella terms.

If the part after the comma is the problem, then surely the part before it is the solution? You need to be looking for net profitability across literally thousands of trades before you can have any real confidence in a strategy. So unless you are trading in an extremely high frequency fashion, there will be no way to accomplish this with paper trading in any reasonable amount of time. Hence you should automate and backtest on historical data rather than paper trading. Paper trading is useful in that it allows you to become familiar with the nuances of your strategy's order entry and execution foibles; it is not a sound way to validate a trading system. Hope that helps.

Hi Tams, Thanks for your reply. Your suggestion is what I had attempted to do (in a very clunky and inelegant fashion!). I found it very difficult to incorporate weekend dates into a filter. I'm not convinced that my solution below works - can you see any obvious issues with it? variables: m(0), count(0), dom(0), sess(0), buywindow(false); sess=CurrentSession(0); If sess<>sess[1] then begin m=month(d); if m<>m[1] then count=1 Else count=count+1; dom=dayofmonth(d); End; If dayofweek(date)<4 then BuyWindow = (m=1 and dom=31) or (m=2 and dom=28) or (m=3 and dom=31) or (m=4 and dom=30) or (m=5 and dom=31) or (m=6 and dom=30) or (m=7 and dom=31) or (m=8 and dom=31) or (m=9 and dom=30) or (m=10 and dom=31) or (m=11 and dom=30) or (m=12 and dom=31); If dayofweek(date)=4 then BuyWindow = ((m=1 and dom=31) or (m=1 and dom=30)) or ((m=2 and dom=28) or (m=2 and dom=27)) or ((m=3 and dom=31) or (m=3 and dom=30)) or ((m=4 and dom=30) or (m=4 and dom=29)) or ((m=5 and dom=31) or (m=5 and dom=30)) or ((m=6 and dom=30) or (m=6 and dom=29)) or ((m=7 and dom=31) or (m=7 and dom=30)) or ((m=8 and dom=31) or (m=8 and dom=30)) or ((m=9 and dom=30) or (m=9 and dom=29)) or ((m=10 and dom=31) or (m=10 and dom=30)) or ((m=11 and dom=30) or (m=11 and dom=29)) or ((m=12 and dom=31) or (m=12 and dom=30)); If dayofweek(date)=5 then BuyWindow = ((m=1 and dom=31) or (m=1 and dom=30) or (m=1 and dom=29)) or ((m=2 and dom=28) or (m=2 and dom=27) or (m=2 and dom=28)) or ((m=3 and dom=31) or (m=3 and dom=30) or (m=3 and dom=29)) or ((m=4 and dom=30) or (m=4 and dom=29) or (m=4 and dom=29)) or ((m=5 and dom=31) or (m=5 and dom=30) or (m=5 and dom=29)) or ((m=6 and dom=30) or (m=6 and dom=29) or (m=6 and dom=28)) or ((m=7 and dom=31) or (m=7 and dom=30) or (m=7 and dom=29)) or ((m=8 and dom=31) or (m=8 and dom=30) or (m=8 and dom=29)) or ((m=9 and dom=30) or (m=9 and dom=29) or (m=8 and dom=28)) or ((m=10 and dom=31) or (m=10 and dom=30) or (m=10 and dom=29)) or ((m=11 and dom=30) or (m=11 and dom=29) or (m=11 and dom=28)) or ((m=12 and dom=31) or (m=12 and dom=30) or (m=12 and dom=29)); If (BuyWindow or DayofMonth(date)=1) then Buy this bar;

I'm hoping that someone can help me out with EL code to identify the last trading day of the month as part of a strategy entry condition? Any help would be greatly appreciated. Bluehorseshoe

Ha! Can't argue with that! Anyway, you've doubtless got better things to be doing with your time . . .

Another amusing trait is that, despite numerous books singing the praises of trend-following, I have watched him conduct webinars where he attempted to scalp crude from a one minute chart. Nevertheless, his books on trend-following are well written and a good introduction to the subject, so I certainly wouldn't dismiss them just because Covel himself is probably a "can't trader".