MAE

["Walter Lake" <wlake@xxxxxxxxx>]

Hi John

Here's the rough work for the KW8U/C8U spread using Moore Research 15 year
data. All of the data is put on graph paper, so that I end up with 15
"hockey sticks" going across the page.

Worst Equity Amount = "Maximum Adverse Excursion" (see Sweeney's "Campaign
Trading" book) clusters at

0 to 3 cents = 4 years
6 to 8 cents = 5 years
14 to 21 cents = 3 years
28 to 33 cents = 2 years
70 cents = 1 year

Best Equity Amount = "Maximum Favourable Excursion" clusters at

0 to 1 = 2 years
6 to 9 = 3 years
12 to 15 = 3 years
26 to 44 = 5 years
60 to 70 = 2 years

Profit clusters at

0 cents = 4 years
3 to 6 = 4 years
13 to 17 = 2 years
26 to 70 (scattered) = 5 years

So this is the raw data.

Am looking for entry criteria, "proven correct" criteria, "MAE" stops (i.e.,
8 cents eliminates most of the nasty moves against) and exit (there's a big
difference between the "MFE" and the profit clusters.

I charted the three variables into the following groups (bins):

1.  MAE (stop) of 0 to 8.5 cents:

a) large MFE & large profit (i.e., only a slight drop off) - years 4, 10,
11, 12, 14 (33%)

b) large MFE but substantially reduced profit - years 1 & 3 (13%)

c) medium MFE & medium profit - year 7 (7%)

d) low MFE & low profit - year 13 (7%)


2.  MAE (stop) 16 to 33 cents:

a) reversal to low profit - years 5, 6 & 9 (20%)

b) continued large loss - year 8 & 15 (13%)


3. MAE very large - went bad and stayed bad - year 2 (? trade this as a
fade, see Ruggiero articles on this)


Now we add the dates

1. MAE of 0 to 8.5 cents (60% of spreads):

a) early MAE, large MFE (late) and large closed profit

Year     MAE         MFE
4         June 24     Aug 25
10 June 29 Aug 23
11 June 24 Aug 25
12 June 23 Aug 4
14 June 27 Aug 20

b) large MFE - then down with substantially reduced profit (must get out
before profit is given back)
Year     MAE         MFE
1 July 16 Aug 8
3 July 10 July 21

c) medium MFE & medium profit
Year     MAE         MFE
7 July 17 Aug 21

d) low MFE & low profit (marginal trades)
Year     MAE         MFE
13 July 17 Aug 16



2.  MAE = 16 to 33 cents

a) questionable trades, i.e., late entry
Year     MAE         MFE
5 July 1 July 17 down, up - 12 cent possible profit
6 Aug 17 July 21 up, down, up - 9 cents
9 July 6 Aug 9 down, up - 12 cents

b) continued large losses all summer
Year     MAE         MFE
8 Aug 23 June 26 - 0 cents
15 Aug 15 June 30 - 6 cents profit



3.  MAE is very large all summer

2 July 12 gone down and stayed down


Hope this helps. MAE can be calculated in different situations using
Metastock, or spreadsheet or by hand on graph paper. I find that spreads are
easier to do by hand. There's only 700 to 800 to look at over the year. What
is billed as 14 profitable years out of 15 turns out to only have 5 or 6
tradeable years. Once again demonstrating that what appears as profitable
may not be tradeable.

Best Wishes to all the guys who have written to me personally. I wish that I
had time to write to you more often. There's so much to talk about.

Walter