I use MP but perhaps not in the classical ay, more of a another way of looking at volume and likely S/R and where we are in a short term cycle. Read Mind over Markets (which deals with MP) for how markets move/auction, CBOT hand book also.
This (originally posted on ET by Cutten) is about the only meaningful thing of length (other than 1 or 2 sentence gems) I have ever read on trading futures. Everything else i.e most books on the subject seems to have been copied from something written on wheat 100 years ago or something for the Oats market in the '70's. Apologies to Cutten for any underlinings/changes added by me.
"I think I have an idea what you may be doing wrong. I'll try to give some concrete trading techniques I think will at least get you to breakeven, and then hopefully you can move on from there.
Firstly, by hitting bids and offers, rather than putting limit orders and waiting to get filled, you are giving up quite a bit of edge. When you hit a bid or offer, there are three possibilities:
i) it goes bid on size - you are now at a scratch, but with a good chance of making a tick
ii) it goes slightly bid, then trades back - you are down half a tick at least
iii) it doesn't go bid at all - you are down at least 1 tick.
So your payoff is 0 or better, -0.5 or worse, and -1 or worse. You are losing 0.5 or more on each trade. So you need to have 0.5 of directional edge just to breakeven, include commissions and you need 0.6 to breakeven.
Compare to waiting to get filled:
i) your offer gets lifted and it goes bid on size - you now have a good chance of making a 1 tick loss, but will sometimes scratch
ii) your offer gets lifted, it goes weak bid - the chance of losing a tick is moderate, chances of scratching if you want are good, chance of a tick profit is moderate (probably similar to chance of a tick loss)
iii) you get filled, and it doesn't even go bid at all, or goes bid and immediately gets offered again - chance of losing a tick is low, chance of making a tick is pretty good, and it's easy to scratch.
So your payoff is a likely 1 tick loss or scratch, a tossup, and a likely scratch or 1 tick profit. After commissions you may need about 0.1 in directional edge to breakeven. Compare a 0.1 transaction cost to a 0.6 cost. A trader who makes 5 trades at 0.1 cost and clears 2 ticks profit per contract per day would be a net loser with a 0.6 cost.
So, buy into selling, and sell into buying. Only enter using limit orders, letting bids trade out to fill you, not lifting the offer or whacking the bid to enter. On exits, if your stop is hit, obviously you do have to whack bids and offers. But try not to do it on entry.
Now, there are *rare* occasions where whacking is right. This is mainly when a trend or meaningful move is taking place. I don't mean a 4 tick move, I mean something that you have a strong belief could result in 6, 7, preferably 10+ ticks move. Then, giving up 0.5+ ticks is worth it in order to participate. But in general, it is better to let the market fill you.
Second - stop using the US contract as a leading indicator. The US going bid is just as likely to be a fakeout as the Bund going bid. Even if it is a *slight* leading indicator, this is not enough of an edge to offset the -0.6 cost of lifting offers and whacking bids. So forget trying to scalp a tick by following the US.
Third - bad entries. One of the greatest causes of losses in ultra-liquid competitive markets is jumping on moves when they are already overextended. Ask yourself how many times you have sold the Bund when it has just sold off 3-4 ticks, or bought it when it's just rallied that much? The guys making money scalping are the ones who are *fading* those trades. For example, one of my rules with normal scalps is to exit at least half my scalp on a 3 tick bounce from the lows or drop from the highs. If the low was 30, I am always offering half at 33, and usually the rest at 34-35 (obviously I don't always get in at the low - maybe I bought 32 and 31, or maybe even 34 and 33). I need a damn good reason to hold on longer. So you need to stop chasing the market. Do not enter on a move after 3-4 ticks unless you have a damn good reason (e.g. if it is a strong momentum market, such as last friday after the payroll report).
Fourth - stops. The Bund, like most liquid markets (e.g. ES), is very efficient at hunting out close stops. You need to place stops not in relation to your entry price, but in relation to where it will be difficult for the market to hit your stop, without that demonstrating that you are almost certainly wrong. You absolutely *must* avoid situations where it is easy for the market to stop you out, and yet you then turn out to be right. This goes with point 3 above - if you sell after a 3-4 tick selloff, then it is very easy for the market to take out a 4 tick stop. If you enter when the market is overextended, and has just risen into resistance, then it is actually quite hard for it to go another 4 ticks without getting some serious selling pressure. If it does go that 4 ticks, chances are you are wrong. Usually if you mistimed your entry a bit, it will go 1-3 ticks, then come back and let you out for a scratch or even small profit.
Fifth - getting a proper entry technique. In normal markets (i.e. non-trending, non-momentum), you want to be buying oversold, and selling overbought. Luckily there is a simple objective method to decide this. Get up some Bollinger bands on a 5 minute chart, use a 10 period moving average, and set the band width to 1.75 standard deviations. Look for buys at or below the lower band, and sells at or above the upper band. Then, look to exit on either a 3-4 tick bounce from the lows, or a move back to the 10 period moving average. You can't do this blindly - a trend will always be overbought or oversold for some time on the bands. But you can adjust for this, given some experience and trial & error. In an uptrend I will wait for it to go 1-3 ticks *above* the band, and exit once it comes back and hits the band or goes a tick or two below. It is very rare for a continual trend which never comes back below the band until you are massively offside. In any case, I have a 4-5 tick worst case stop on these fade trades - it gets hit maybe 10-20% of the time.
This indicator is also good for avoiding crap entries - do not enter in the direction of momentum when the market is at or beyond the bands, unless you have a damn good reason, as the odds are generally against you. If you are short and the lower band gets hit, start thinking seriously about why you are not covering. Generally I always cover at least half when it hits the lower band (and vice versa for longs on the upper band).
Next point - trends, volatility breakouts, and momentum markets. You know the type - the days when the market just keeps going, or is moving wildly. Forget oversold/overbought here, you need momentum techniques. A very simple one - wait for the direction of momentum to become clear, then with faster markets enter on breakouts, and with slower trends, enter on 40-60% pullbacks, preferably to previous intraday support/resistance. Place a stop below the low of the most recent 5 minute bar in slower trends, and in busy markets, use a shorter timeframe e.g. 1-2 minute bars. This method can often capture very nice moves, when you get on them. The idea is to lose 4-5 ticks when wrong, and make 10+ when right. Look at the action on friday after the payrolls, to get an idea how well this method can work - after the initial couple of minutes wild action, the market dipped to 112.26-27, then started a powerful rally (you could see the momentum much better on the day, using market depth, than you can now from the 1 minute chart). I bought 84s and 85s on the break of the previous 1 minute bar high at 82, and sold at 91 and 92 (near the previous high - too early, I admit). I was then thinking ok, where do I get short? Rather than try to top-pick, I waited for the 1 minute bar low to get broken - I shorted at 00 and 99, then covered for a scratch as it broke above the previous 1 min bar high. I then reshorted at 99, 98 and 96 as it broke back below the 1 minute low, giving the same signal. Using a trailing stop of the previous 1 minute bar's high, I was able to run the position quite some way. As it happens, I covered at 84, using discretion as I felt it was at a give-up point for longs with that big down bar, but 87-88 would have been the "system" exit. Obviously I had losing trades that day too, but overall I was up a reasonable amount and did ok.
Next step - levels. Odds are improved when you use support and resistance levels. These are clear - previous day's high, low, and close. Major highs and lows from the daily chart. And, to a lesser extent, areas of intraday support and resistance. The ideal setup is an overextended market, at or above the upper Bollinger band (or vice versa for downmoves), also running into a resistance level. Go short, average in once at 1-2 ticks higher, use a 3-4 tick stop from your 2nd entry price. Cover half at 3-4 ticks back from the high, move stop to a tick above the high on the rest, exit the rest on a return to the 10 period moving average (or earlier if it looks like it's not going down). That's my bread and butter trade, where I have a good (60-70%) win rate, and often make profits equal to my stop and sometimes more.
cont.......
