Odds Czar: Simple Biases in the Futures Markets 2005

setharb wrote

no, they're just static daily indicators. you had an opportunity to take profit intraday (not on any indicators i was presenting admittedly) --you would have certainly had pain had you held them all day. but that's just one for the "L" column.

oct 20 suggests another sell. maybe this one will fare better

 

The fourth component of our daily biases is sort of a compound idea. Or maybe it’s actually just a single concept affirming something my partner is always voicing. If a signal isn’t strong enough to meet a buy criterion or it stops meeting the condition, thereby signaling a sell exit...

...is it actually qualifying as a reverse signal short sale? Because after all, if something isn’t strong enough to be a buy bias, it has to be a sell one, however slightly, right?

Obviously, when you get to actual systems, there are middle grounds where it’s not strong or weak enough to justify any move. Right now, we just want to know which way the meter teters.

If a daily range is smaller than the average 10 day range, the direction of the close, up or down, suggests the next day’s bias. In other words, if the range is smaller than average, buy after a higher close or sell after a lower one. Table 1 reveals 5 year results in our targeted financials.

Accept on faith that acting off smaller ranges works better than always going in the direction of the previous close. (It does). By definition then, higher ranges would have to be generating opposite results—buy following down closes, sell short following up ones. Right?

Right! Table 2 confirms that. We now have to halves of a pretty singular direct concept. The combined halves of table 3 is our actual fourth either-or bias indicator.

Table 1
Net Profit # Trades % Profit Avg Trade
S&P (full) $54,037 668 51.5 $80.89
Nasdaq (full) $49,700 661 50.23 $75.19
Russell (full) ($25,425) 645 49.46 ($39.42)
30 year Bond $16,062 688 50.73 $23.35
10 year note $1,859 702 47.58 $2.65
5 year note $2,250 678 48.08 $3.32
Japanese Yen $5,675 703 51.07 $8.07
Euro Currency $7,950 693 49.35 $11.47
Swiss Franc $14,900 703 50.78 $21.19

Table 2
Net Profit # Trades % Profit Avg Trade
S&P (full) $140,875 525 54.67 $268.33
Nasdaq (full) $53,800 528 51.7 $101.89
Russell (full) $52,875 550 51.45 $96.14
30 year Bond $468 477 49.27 $0.98
10 year note ($10,171) 477 48.85 ($21.32)
5 year note ($5,657) 473 49.26 ($11.96)
Japanese Yen $13,475 474 49.79 $28.43
Euro Currency $27,100 489 54.19 $55.42
Swiss Franc $10,275 479 52.61 $21.45

Table 3
Net Profit # Trades % Profit Avg Trade
S&P (full) $194,912 1193 52.89 $163.38
Nasdaq (full) $103,500 1189 50.88 $87.05
Russell (full) $27,450 1195 50.38 $22.97
30 year Bond $16,531 1165 50.13 $14.19
10 year note ($8,313) 1179 48.09 ($7.05)
5 year note ($3,407) 1151 48.57 ($2.96)
Japanese Yen $19,150 1177 50.55 $16.27
Euro Currency $35,050 1182 51.35 $29.65
Swiss Franc $25,175 1182 51.52 $21.30

 

Here are my futures biases for tomorrow, October 21

A "1" means bullish bias. A "-1" means bearish bias. The total is the sum of biases. A positive sum will be long (L) bias. A negative sum will be a short (S) bias. A sum of zero will be a neutral (0) bias.

 

Art,

I might deal with a shorter time frame then yourself, but I disagree with you to a point. I have spent the last six years scalping interest rate futures and believe that there are actually three (3) different stances toward the market. Buy, sell, and wait.

It has helped me get rid of the, "Well if it's not going up it's gotta be going down!" This mentality cost me dearly in my first 4.5 years of trading.

If i am biased long and that signal has stopped, I like the long signal to be reconfirmed or some type of level to be broken to go short. And vice versa. If not, I feel like I am trying to call a top or bottom. Tops and bottoms need to be confirmed no matter how much you think it might cost you. Embracing this idea has made my trading much less stressful.

Great thread.

 

buylo wrote

you raise some vailid points buylo. thanks.

 

I don't think good discretionary traders avoid the hard trades. That is a matter of good discipline and not related to whether one is a discretionary or mechnical trader.

Trading is part art, part science. With a mechanical approach it assumes it is only science. Where does the judgement come in? Also, you can't code and backtest every strategy that has a set of rules. For example, if candlesticks are part of methodology it is better to manual test them as not everything can be coded.

Trading isn't just about rules, it is about getting inside the head of other traders and figuring out what they are feeling about their trades. The rules are merely an expression of capitalising on that. Most of all trading is about reading the market and thinking about what the market is telling you. That is why no mechanical system will ever beat a good discretionary trader.

 

LondonUsTrader wrote

a good discretionary trader is the king of the hill--no question. mechanical traders in my book have acknowledged that. again though, how many of us are born that way? that's pretty much the reality--you're born into that rarified elite or you're not. i know many traders, and maybe enough successful "feel traders" to count on two fingers.
the many things we CAN teach ourselves almost inevitably could be reduced to rules. if it can be a rule, it can be codified. candlesticks are just another visual way of depicting standard open high low close bar charts.
for maybe 99.99 percent of us, trying to "hear what the market is telling us" is futile, giant industry lore to the contrary. for all practical purposes, the market ain't "saying" squat.

 

We return to ultra-simplicity personified with the fifth either-or indicator. Instead of comparing a close to closing averages, we key it off high-low averages—in this case, incorporating 15 days. ((average(high,15))+(average(low,15))/2). This slight variation produces some interesting results as table 1 confirms.

Net Profit # Trades % Profit Avg Trade
S&P (full) $87,612 1204 49.83 72.77
Nasdaq (full) $144,950 1202 51.66 $120.59
Russell (full) $27,350 1202 49.33 $114.24
30 year Bond $23,343 1198 48.91 $22.75
10 year note $11,437 1198 50.25 $9.55
5 year note $13,312 1200 50.67 $11.09
Japanese Yen $5,487 1198 51.34 $4.58
Euro Currency ($19,388) 1197 48.79 ($16.20)
Swiss Franc ($12,038) 1199 48.71 ($10.04)

 

A 15 minute audio interview of me is begin rerun this weekend. I explain many of my trading philosophies, and what goes into building good systems. It's featured on the home page.

http://www.tigersharktrading.com/

 

Just a word of caution. Above will not work for daytrading Indexes. Especially if open gaps.