Caution: Math Geek at Work

I don't know if he's on now, but I'm pretty sure he just means that the domain of the function is the Reals (as opposed to some subset like the Integers or the Rationals).

Interesting...I think the only thing we can know about price is when it has reversed, within certain descriptive limits (statistical profile), and where it is likely to reverse again, within the same set of descriptive limits. All taken with a grain of salt. Keeping one's day job...

Thanks Scott, Thought that would be understood.

LOL, you're right it woud be. : I think it's kind of a habitual common practice to specify the domain of a function whenever mentioning it for the first time. If you don't hear something like "over the reals in the closed interval [-1,+1]" after the description it just sounds "wrong."

This past week was finals week at university... I guess I'm still thinking in engineering mode . I also have a thing for specificity... you can't support or refute a statement without knowing what it actually says .

I'm with you there. You'll notice in the later research projects on my site that I always have a "Dataset" section in the notes that describes exactly what data I used for the analysis and where I got it. Down with vagueness!

No chart today. Not much to report. The price did not hit my stop and it has recovered, so I'm still long 52 units at about B/E. We now have an upleg in progress but it hasn't reached the exhaustion level, so I'm not placing any sell orders yet. That is all. :

Hola!


I'm making another refinement to my position management method. I'll illustrate it using the current EUR/JPY position.

The price continued up past the exhaustion level, so I was cleared to start placing the buy orders below it. So I started at the last local bottom as usual and started placing them up from there 50 pips apart, in larger numbers of units 1,2,3,4...etc. Then I calculated how many sell orders I would have to place before the average proceeds from them would be above the average cost of my current long. It turned out that I would need to get to the 12th order for this to happen.

Ok, so THEN I thought to myself... (wait for it) ... "SELF," I thought, "this is sort of dumb. If the point is to place sell orders at relatively high prices, then why the HELL am I placing any at ALL down near the last trough?"

The reason I use the last peak/trough of course it that it gives me a point from which to start measuring the length of the current leg. Then as the leg gets longer and longer, the size of the orders I'm placing gets larger in anticipation of the reversal. So in the case of the EUR/JPY position, I would have been up to the eighth order by now with a size of eight units. This would be a total of 36 units worth of short orders, Σ(1,2,3...8).

Well ok, so if I'm supposed to have 36 units worth of sell orders in, then why not "collapse" them all into one sell order which is at the highest price? Above that point, I can continue to place sell orders as usual every 50 pips, 9,10,11, etc. The best way to apply this rule is to wait not only for price to pass through an exhaustion zone, but to be on the profitable side of the average position price as it is now.

The disadvantage to this refinement comes when the price only makes a small retracement and then continues on its merry way upward. Using the old method, only a few of the sell orders would be hit and I would be long a larger amount for the continuation. But this sort of objection is based on the idea of "trend continuation" being more probable than a full reversal. I don't believe that is generally the case, so I'm going with the refinement.

Here's the current situation with the new tactic:



But most of these orders would have been way below my average cost.

I just tried another intraday session, and ended up at about a dollar below breakeven again. The only value I got out of the session was a better understanding of the dynamics of my position management system. After seeing it work in a short time frame, I have even more confidence in the concept.

However, using it in an intraday session just doesn't work because the method can't have a "deadline" for closing out the position. The method only works when the position can remain open indefinitely, and I can make adjustments at each bar. But with an intraday session, I have to eventually shut down to eat, sleep, or whatever. So if the position isn't in a profit by that point, I have to shut it down for a loss.

Therefore, I'm going to bag the intraday trading and just continue with my two long term position management methods (Lichello's AIM with NZD/USD and the "trailing order" method that I've been using with EUR/JPY). Those two positions take about 10 minutes a day to manage.

As it is journaled.... so it is done.

Looking good! My open equity is currently up 6.7% from the inception of this method. I have a few weeks worth of EOW data, so I'll post that this weekend.

After collapsing the first 8 sell orders into one at 155.50 (described in an earlier post), I placed the 9-unit one yesterday at 156.00. That one had already filled today but price had gone back up enough for me to place the 10-unit order at 156.50. So since I was long 52 units before and have sold 9, I'm now long 43 units. The open sell orders now total 46 units, so if they all filled I would be short 3.

All my stops at 147.00 are cancelled since they're covered by open sell orders. The stops for these sells is up at 165.00.

The whole account was up 3.2% this week to just under 80% of the original balance. I continue to use very small position sizes as the account equity moves sideways. I will increase size only when the equity curve is clearly heading north.

I haven't posted the total account equity chart in a long time, mainly because it still remains flat and boring, but here it is:

As promised, here is a chart showing the open equity of just the subaccount where I hold the EUR/JPY position. I keep it separate so I can track the equity curve of just that method. This chart shows the 5 weeks since I started keeping track.

Here's another refinement to my "trailing order" position management method. This is in keeping with the basic idea of setting the sell orders as high as possible and the buy orders as low as possible. The refinement is simply that if orders are filled causing gaps in the sequence, then I'll move the farther orders forward to fill the gap, thus raising my average sell points and lowering my average buy points. The current situation illustrates the idea.

As EUR/JPY has been rising, it has also been choppy lately. I've been placing sell orders behind the price as usual, and have gotten up to the 13 unit order. However, some of the other orders have already filled due to the chop. My actual long position is now down to 22 units.

Anyway, I still had that big 36 unit order (consolidation of orders 1 through 8) open, but the 9,10, and 11 unit orders filled. Price has gotten high enough to put the 12 and 13 unit orders in. So what I had is this big 200 pip gap in my sell orders. It looked like this:



There's no reason to have a big 36 unit order sitting way down below the other ones like that, since the point is to sell HIGH. So I got rid of that order and replaced it with a set of orders in the normal sequence. So instead of having the orders go 36...gap...gap...gap...12...13, they now go 8,9,10,11,12,13. I basically just moved the orders up to fill in the gap. The 8,9,10 and 11 orders actually add up to 38 units, but that's fine. So here's the new situation.



As I make each little refinement to the basic method I get more and more confident in it. Wow, could this be the HOL.... errr, never mind. :

No charts today because it's a simple update. The continuing upleg of EUR/JPY has reached above 160.50 and I have placed additional sell orders for 14, 15 and 16 units behind it, with the last order being at 159.50. My long 22 unit position is now about 420 pips in profit.

My sell orders are in a sequence from 8 units to 16 units for a total of 108 units. If they all filled I would be short 86 units.

Another short post. The sell order for 16 units filled on a dip, so now I'm only long 6 units. I think my long position on this upleg got as high as 52 units (?) at one point, so it's almost all sold out now. Let's see how good the timing was. If price turns south right here that would be epic.

The overall account was down 0.5% this week on NZD weakness, but the EUR/JPY portion continued to gain. Here's the updated chart of the open equity for just that portion of the account.

EUR/JPY continued up today, I'm still long 6 units, and I've placed the 17-unit sell order at 160.00. Because the 16-unit order at 159.50 already filled, there was a gap between my 15 and 17 unit orders. So conforming to my policy of pushing the average potential average sell price as high as possible, I cancelled the two lowest price orders for 8 and 9 units and replaced them with a single 16-unit order at 159.50, which is where it was before.

Using the "Art of War" lingo from fti's thread, I moved the two soldiers in the rear of the column forward to plug up the gap near the front.

So now I have a nice uniform set of sell orders starting at 156.50 up to 160.00 in the sequence 10,11,12,13,14,15,16,17.

Every time a gap appears in the sequence due to a previous fill followed by a continuation of the current leg, I'll be doing the same "gap filling" routine. It's just a way of pushing the orders in the most advantageous direction (up for sells and down for buys).

I should have done this before the last post. Anyway, my criteria for placing an order at the next 50 pip mark is that the price has moved at least 100 pips beyond that point. Since EUR/JPY had reached a high of 161.72 earlier today, I'm clear to place another sell order at 160.50.

So I now have an 18 unit sell order in at 160.50.

This upleg is the longest leg I've seen since starting this method. I have a total of 126 units worth of sell orders lined up behind it now.

Aaack. That 18-unit order filled, so the EUR/JPY position has now flipped to net short 12 units. Could that be drawdown country I see up ahead? :

Haven't posted a chart for a while so here's where we are now: