what happened to george?

George: Cant answer private messages as I am a freebie. I answered your post in my previous one to Zeev.

GEORGE,

It looks like the countertrend day will be on Thursday

George,

The question is stay long or go to sidelines?We are getting overbought on the daily

Regards

Marc

George,

We need to hold 1335 at the close or I will go to cash.

Regards

Marc

I think this correction is over

seems the market is one place that rewards solitude. and boy are alone in making that call!

George,

The VIX closed above it's upper BB only the fourth time this year today.The total p/c went to 1 and the TRIN closed above 2.0.I think this correction is over.

Regards

Marc

George,

Now for the third time.This may be the day to buy.We will see.

Regards

Marc

George,

We would have to go down to 1360 or so before I would think about buying.

Regards

Marc

George,

Now this day may be the day to buy at the close.We will see

Regards

Marc

George,

This may be the day to buy at the close after all.

Regards

Marc

George,

My charts tell me we could have 1 more day up but with Naz at 7 days up in a row I will go to cash and wait at the close.

If we do go up tomorrow I will probably short.

Regards

Marc

George,

If we close strong I will go to cash to wait out a suspected countertrend day tomorrow.

Regards

Marc

George,

I see the next countertrend down day as Wednesday.Today we should close as high of year.

Regards

Marc

George ,

Are you getting that from stockcharts?I have to join.What happened on the 6th day?


Regards

Marc

2/1995

George,

Do you know when the last time the SOX closed 5 days above the upper BB?

Regards


Marc

I have a set of rules that I follow(several of them quite counter-intuitive)

then they're probably right. -g-

>>>>>>>>>>>>>>>>
I wonder if you still look at QQQ historical intraday patterns to make your trading decisions. Also, what does your position size depend on?
<<<<<<<<<<<<<<<<

No. looking at intraday patters left out too much information, and seemed to work fine in a range-bound market, but badly in a trending one.

The new system trades in the direction of the current trend, but it's overruled by OB/OS situations to try, and anticipate, a change in trend. The trick, of course, is in the determination of the current trend, and when to consider a situation as overbought/oversold, and when to disregard it...

All of the above is automatic. Position size is still determined by myself. I have a set of rules that I follow(several of them quite counter-intuitive) but they are still evolving. Some day I might publish something about how to size Dynamic Funds positions, but I need to gather my thoughts together first. <g>

The main problem which these systems is that they are good for several months, but then they implode. (Taking us down with them. <ng>) My previous system served me well for more than a year, but has been useless since last April. (I keep tracking it, just in case...)

H.

Hi Hul, I appreciate your very interesting info. I wonder if you still look at QQQ historical intraday patterns to make your trading decisions. Also, what does your position size depend on?

I haven't recently found too many traders on the threads who can consistently gain even 6% a month (100% a year) for more than a year (of course many did it easily during the 10/1998-3/2000 period). The best traders with at least a partially verified record that I encountered recently are:

(1) NDX funds/QQQ: ajtj (Profunds), Huluriasquias (Rydex), positiontrader (Profunds), wahz (QQQ)

(2) Options/futures: ajtj (options), TradeHard (futures), WinLoseOrDraw (options, futures)

But even among those traders above, very few if any have a record of average compounded 10%+ returns for the past 12 months (even though they may have achieved it for shorter periods of time). I also saw a few other traders on SI claiming 10%+ a month returns in the past and the actual record in one instance verifying the claim.

exp,

the question is still open.

After I published my results here I had a couple of flat months, and a -5% one. That wouldn't be too bad, except for the fact that I have to take money out regularly for living expenses.(And my capitalization is not too great to say the least...)

So I put myself to work, and I came out with a completely different system that I started using 6/10/03. In back-testing it yielded more than 15% a month, but in real money, "forward-testing," it has yielded around 10% per month in each of the first 2 months, and it's flat for the first third of the third month so far.

I'll be very happy if any of these systems could stand the test of time, and give consistent results for at least over a year. (I'm trying... <g>)

What follows is a "real-account" transcript of the system. This is done with Rydex Dynamic funds (Velocity, and Venture) trading only at the end of the day, and not always being 100% invested. (Usually between 70%, and 95%)

(I apologize for leaving the dollar amount in --I know it's tacky-- but I didn't know how to remove it without affecting the formulas...)

Note: I'm long for today 8/22.

Date, Amount, Daily, From 6/10

6/10/03 $23,124.24 0.00% 0.00%
6/11/03 $23,559.26 1.88% 1.88%
6/12/03 $23,598.91 0.17% 2.05%
6/13/03 $22,741.51 -3.63% -1.66%
6/16/03 $24,149.82 6.19% 4.44%
6/17/03 $24,081.92 -0.28% 4.14%
6/18/03 $24,175.22 0.39% 4.54%
6/19/03 $23,332.83 -3.48% 0.90%
6/20/03 $23,245.50 -0.37% 0.52%
6/23/03 $23,966.37 3.10% 3.64%
6/24/03 $24,194.77 0.95% 4.63%
6/25/03 $24,319.06 0.51% 5.17%
6/26/03 $25,265.94 3.89% 9.26%
6/27/03 $24,983.22 -1.12% 8.04%
6/30/03 $25,088.90 0.42% 8.50%
7/1/03 $24,493.65 -2.37% 5.92%
7/2/03 $25,575.57 4.42% 10.60%
7/3/03 $25,003.04 -2.24% 8.12%
7/7/03 $26,978.86 7.90% 16.67%
7/8/03 $26,358.41 -2.30% 13.99%
7/9/03 $26,505.55 0.56% 14.62%
7/10/03 $25,508.56 -3.76% 10.31%
7/11/03 $26,014.14 1.98% 12.50%
7/14/03 $26,652.75 2.45% 15.26%
7/15/03 $26,617.30 -0.13% 15.11%
7/16/03 $26,705.36 0.33% 15.49%
7/17/03 $28,211.51 5.64% 22.00%
7/18/03 $27,986.42 -0.80% 21.03%
7/21/03 $28,666.18 2.43% 23.97%
7/22/03 $27,958.64 -2.47% 20.91%
7/23/03 $27,561.34 -1.42% 19.19%
7/24/03 $28,128.22 2.06% 21.64%
7/25/03 $27,125.62 -3.56% 17.30%
7/28/03 $27,257.77 0.49% 17.88%
7/29/03 $26,987.40 -0.99% 16.71%
7/30/03 $26,569.37 -1.55% 14.90%
7/31/03 $26,203.86 -1.38% 13.32%
8/1/03 $26,556.42 1.35% 14.84%
8/4/03 $26,427.14 -0.49% 14.28%
8/5/03 $27,751.30 5.01% 20.01%
8/6/03 $28,219.51 1.69% 22.03%
8/7/03 $28,303.57 0.30% 22.40%
8/8/03 $27,828.11 -1.68% 20.34%
8/11/03 $28,420.57 2.13% 22.90%
8/12/03 $27,747.21 -2.37% 19.99%
8/13/03 $27,748.21 0.00% 20.00%
8/14/03 $27,782.31 0.12% 20.14%
8/15/03 $27,414.80 -1.32% 18.55%
8/18/03 $28,505.88 3.98% 23.27%
8/19/03 $27,954.91 -1.93% 20.89%
8/20/03 $27,997.36 0.15% 21.07%
8/21/03 $28,505.71 1.82% 23.27%

NoiseBox includes intelligent position sizing plus it is intraday which enhances the probability of profitable trades. So, it should be more profitable than a similar system based on eod closes simply because of higher frequency of trades.

On the other hand, Huluriasquias claimed returns of 7-8% a month compounded for the past 7 months earlier this year using Rydex funds (I believe) and eod trades and positiontrader (Marc) has had returns exceeding 13% a month compounded for the past 10 months using Profunds and eod trades.

So, it is still not clear to me whether 10-12+% a month is achievable and whether eod trading is able to achieve it. Of course, even 6% a month means 100% a year which is surely a superior return.

the NoiseBox is now 100% mechanical. or it would be, if i didn't have an ego and a need to retain some sense of self-importance by not turning it on. -g- i'm running it in parallel to see how much "edge" my discretion is adding.

so far i'm not very impressed with myself, lol.

Mechanical systems are too simplistic in most cases:
(1) There are multiple factors that need to considered in trading decisions with appropriate dynamic weightings.
(2) Markets change over time so systems based on data mining over a restricted time period are of limited reliability.
(3) Position sizing may not be considered in many mechanical systems.
(4) Diversification may not be considered in many mechanical systems.

thought you might find this datapoint intersting, post made by someone else, somewhere else...

08-21-03 02:25 PM

Quote from Funster: And your return per year?

LOL...

46.54% Annual Return...

Max Drawdown 1.4%...

...

So what's really your point in asking this... It's not for sale...

But here's one interesting statistics...

I tested a bunch of optimizable systems starting with specific but robust non-linear criteria included under 5 min. bar, closing the position at the end of the day...

Total number of tests were: 124416

Out of 124416 systems in 3 years of testing data: 1848 were profitable (1.4%)

Take the 1848 and put that in 5 years of ES with Sharpe Ratio of over 1.0 (unfortunately that the max number of years I have in ES data):136 (7.3% of 1848)

Put that into 1987-current SP data: 26 systems with Sharpe Ratio over 1.0 (19% from 5 years)

Out of all the tests... I only got 3 systems with Sharpe Ratio above 1.5 and annual return of over 30%

I don't know the answer either but I will certainly try to find out through my trading. Profunds funds are an ideal tool for that IMO. The goal in any case is to approach the theoretical upper performance limit subject to acceptable risk level (I still estimate this limit at 20% using Profunds 2x funds). I might consider the risk level acceptable if, say, maximum drawdowns do not exceed 10-15% of the total account value. I have to see in practice if this risk level is achievable and compatible with high returns.

Surely risk cannot be eliminated but it can be minimized/optimizedd

i don't know the answer, but i strongly suspect that minimized/optimized risk is incompatible with superlative returns. and 10% a month is in the upper echolons of superlative, even.

WLD, I am reading several books dealing with issues of market risk, bubbles, trading, etc. as well as the academic literature on market inefficiencies. They are all very informative as they deal with the pertinent issues of trading systems, position sizing/leverage, and also emotional/psychology aspects.

Surely risk cannot be eliminated but it can be minimized/optimized. While my preferred approach is basically momentum investing using leverage, hedging and diversification are two techniques that I may use more extensively in the future to manage risk.

I am reading now Lowenstein's book on Long Term Capital Management

are you enjoying it? i found it not only informative, but very entertaining as well. i hope he writes a book about Enron.

it is difficult to earn great reward without taking great risk. but taking great risk means occassionally losing big. interesting problem. -g-

Lisa, I simply meant that, at best, T/A predicts the up or down market moves with a high probability but not certainty. For example, according to T/A (not to mention F/A) the market can be extremely oversold and yet still keep going down (Sep 2001) and it can likewise be extremely overbought and still keep going up (late 1999). Or, there can be a single day with an extreme market movement down (Oct 1987) or up (Jan 2001).

So, it is not optimal to bet 100% of the account based on T/A predictions since there is always a non-negligible probability that the oversold/overbought conditions will be continue. That's why money management (i.e., positon sizing) is the crucial aspect of a successful system. This is explicitly embedded in the Turtle system of Richard Dennis (who earned $200,000,000 through trading) and also emphasized by Ed Seykota (who had 100% average annual returns for 12 consecutive years and developed the first computerized trading system for a major firm).

Now, there are some tools in probability theory that are of help in estimating the correct position sizes (Kelly's formula for the fixed-fraction betting system) and in calculating the chances of "blowing up" that is losing 100% of the account (solution formula for the gambler's ruin problem). These are a bit too simplistic for the stock market case but still provide important insights.

All this becomes even more important when using leverage (margin, Profunds 2x leveraged funds, options, futures). Leverage is absolutely crucial in order to achieve average returns of, say, 100+% a year as many top traders have done (trading futures, options, commodities, etc.) And yet, without disciplined money management, leverage exacerbates that much more the risk of "blowing up" the account. Incidentally, I am reading now Lowenstein's book on Long Term Capital Management where some these issues are brought up.

"T/A unfortunately does not provide satisfactory answers to many questions regarding system signals to make them very reliable."

I'm not sure what you're saying in that sentence, George.

Could you give us several examples?

ATTENTION: Modified System will be used for trading from today.

(1) As far as the details of the Modified System, it will soon be evident from my trades.
Generally speaking, there will be more trades in the direction of the trend and fewer against the trend.
Also, modified position sizing will be utilized.
Finally, trading vehicles will be more diversified.

(2) Modified System might be considered an euphemism for Malfunctioning System. However, the reason for the change is that I have carefully studied what conditions my system must necessarily satisfy in order to achieve 10+% monthly returns.

The first one is: avoid lost opportunities connected to large market moves (in either direction).
The second one is: avoid excessive losses.
The third one is: limit position risk to avoid either being stopped out of positions or incurring large losses.

(3) The Modified System is still not mechanical as I have not found satisfactory mechanical rules to apply in order to achieve optimal entries and exits.

(4) T/A unfortunately does not provide satisfactory answers to many questions regarding system signals to make them very reliable. So, as Ed Seykota, Richard Dennis, and others have noted, the answer must lie in optimizing position sizing/money management aspects of the system.

George

George

I went back to cash

George,

I stayed short.I am looking forward to the down day so I could switch to UOPIX.The SOX is defying gravity

Regards

Marc

George ,

Unless things change the new trend is upward. This will provide a good shorting opportunity for one day then back long.

Your thoughts

Marc

10 trades a month at 60-40 is only 2 net trades to the good. seems doable, look forward to seeing this develop.

For the period 1/1/2000-8/14/2003 which is 43.5 months, the average monthly sum of absolute daily NDX % changes is 49.11 which I rounded to 50% a month. However, for the period 7/1/2003-8/14/2003 which 1.5 months, this sum is only 36.5% or 24.3% a month which is roughly half of 50%. Given this, the expected returns from trading during the past 1.5 months might be proportionately lower due to this lower volatility.

As far my lack of trades for one week, a few points. First, that 2% daily NDX change (actually 2.36% a day for 2000-2003) is not uniformly distributed so gains are larger when NDX turns from very overbought or very oversold. Second, the correct trade was long entry late last week and early this week and less so now. Third, after I exited longs late last week due to lack of clarity on my system signals I spent a few days reevaluating my system and developed substantial improvements I mentioned earlier.

To reiterate, there is no need to trade every day. It is sufficient to trade when either the win probability or the potential gain is the highest. So, say, 10 trades a month timed for the most suitable setups (i.e., most reliable signals) will likely provide the great majority of theoretical gains possible in a given month.

ok, i get the 10% unleveraged, we're coming at it from wildly different directions, but that's cool, it's the same result.

one thing i'm wondering, then, is why you haven't had a trade in a week. using your 60-40 and other assumptions, you have to trade every single day to get that 10% unleveraged return. 50% over 22 days is roughly 2% a day, so every 5 days of inactivity you miss 3 wins and 2 losses, or, net, 2% of your monthly target.

put another way, if you only did 5 daily trades in a month, your expectation would only be 2% unleveraged.

WLD, let me explain the 10% number...

assume that we have these daily % changes in NDX during the 10 trading days next month:

 
 +2% -2% +3% -2% +1% +5% -2% +1% -3% +4%   
 

 
 +2% +2% +3% +2% +1% +5% +2% +1% +3% +4%   
 

I think what I meant is that when I misjudge the trend direction I should lower my exposure and, conversely, when I judge correctly the trend direction I should increase my exposure...I do not have an algorithm for position sizing only a general concept...the broad principle is to cut your losses short and ride your winners applied to the overall account performance...

to counter the gambler's ruin problem I am attempting to dynamically decrease my position sizes after unsuccessful trades and increase them after successful trades

doesn't that guarantee you'll have the least money on the table when things start going "right" and the most money on the table at the very instant they start going "wrong"?

lots of "%" in that post. is this summary correct? your goal is average 10% of the daily trading $ndx range over an extended period of time. so if the sum of all the (high-low) for july was 1000, you'd want to have captured 100 $ndx points, net?

if the range you were talking about was close-to-close, i'd say that's attainable, but i don't think capturing such a large chunk of the intraday range will be possible without going to intraday trading.

it is easy to measure how much potential gain is being given up restricting to EOD by getting a long-run ratio of candlestick shadow length to body length. the ratio is high enough to cause palpitations in any proponent of efficient/equilibrium-trending markets. -ggg-

everything i'm seeing is pointing to EOD as a real discontinuity, and you can see this by the relative infrequency of pivots happening on day boundaries. based on sheer randomness we should see them more often than we do, about 4 times as often as we do by my calculations (which i freely admit could be very very off).

but i've been plenty wrong before, and since being wrong would mean i'd learn something very important, i hope i'm wrong now! :)

Lisa, I like your comments and I know you are a very sharp person. You are right in everything you say. Here are my thoughts.

(1) There are three aspects of trading that have to be mastered to achieve high returns: system development, position/money management, and emotional discipline.

I am working on all of them simultaneously. You can see in my trades that I vary position sizes and I also vary the trading vehicles (NDX, SPX, US Bonds). I am very mindful of gambler's ruin problem which is more acute when trading leveraged funds like Profunds. That is also why the trades are based on a system which is not fully mechanical. In addition, to counter the gambler's ruin problem I am attempting to dynamically decrease my position sizes after unsuccessful trades and increase them after successful trades.

The emotional discipline is maintained by entering trades based on valid system signals and also by pausing trading and re-evaluating the system when the results are sub-par. Positive trading results increase my confidence in the system and therefore my emotional capability to trade based on system signals. Position sizing also plays a role in maintaining emotional discipline since lower position sizes reduce my risk after unsuccessful trades.

(2) In general, based on my readings and observations, I believe that markets are largely efficient so that above-average returns (adjusted for risk) are very difficult to achieve.

Yet, a number of individuals were able to achieve above-average returns historically, possibly due to "pockets of inefficiency" in markets. I am trying to determine through data analysis if above-average returns can be achieved and to what extent (say, 10% a month on NDX on average using EOD trading).

(3) My comment regarding institutionalizing my system needs to be further elucidated here because of possible (unintended) mercantilistic connotations.

The only interest I have as far as "institutionalizing" my system is to do what some of the most famous traders have done in the past. I intend to trade, if successful, until I grow my capital to a reasonably large size. At that point, and only at that point, I would consider starting a hedge fund of which my own money would constitute a large percentage. The idea is to continue trading using my system as usual for years but, at some point, accept large accounts for pooling with my own account in a hedge fund format.

In fact, out of curiosity at this point, I am reading up on hedge funds performance and strategies to assess performance levels and consistency expected in the hedge fund area. Interestingly, hedge funds make substantial efforts to develop proprietary trading strategies which have academic validity in terms of market anomalies (departures from market efficiency). Financial Analysts Journal is an accessible source of information on these issues.

My own experience over a number of years suggests I'm better at analysis of the market, discerning potentially attractive and profitable systems (so to speak) than implementing those systems. In other words, I find it very difficult to put into actual practice, day after day, what I believe to be a profitable approach to trading.

It all looks so easy when I review the charts at the end of the day and do my postmortems.

The human factors in trading, the psychological side of the game, are very poorly understood and clearly interfere in my experience with running the programs that look so doable after hours.

I personally do not believe anyone can do say 10 to 20 per cent per month, month after month, year after year.

Nearly all who believe that's possible either blow up or go on to write books or give seminars where they teach others to make "extraordinary gains" in the markets.

Why don't they spend their time trading and compounding their gains? I suspect it's more profitable to teach others to do it for $29.95 to $5,000 per seminar or maybe write a $50 book.

Marc, in general I prefer high probability trades and right now I don't see any obvious ones.

Frankly, I am actually more and more reluctant to publicize my trades. On the one hand, when I publicize my trades I gain feedback from others and, more importantly, I feel peer pressure to do very well since my trades are public. However, I also believe that my system has a high potential to deliver exceptional returns. And so, I would like to maintain the proprietary nature of my system since I may institutionalize it down the road. That's why it is very possible that I will have to discontinue publicizing my trades after one or at most two additional months of 10%+ returns (if that happens, of course).

WLD, as I stated before, my goal is to capture the maximum possible percentage of that 50% average total NDX daily % changes per month. I am constantly analyzing NDX data to see what percentage of that 50% monthly total is achievable on average (consistently in the long run). It is really mainly a question of what's theoretically possible given the probabilistic behavior of daily NDX changes. An approximate answer to that question should imply a corresponding actual monthly trading return.

My conjecture is that capturing 10% a month of NDX daily changes is the theoretical limit in the long run. This 10% a month is equivalent to a net 20% of 50%
(simplifying, assume 60% of wins and 40% of losses, so
expected value= .60(50%) + .40(-50%) = 10%).

Capturing 10% a month in NDX translates into capturing 20% a month in Profunds (ignoring compounding effects). 20% a month equals about 9-fold increase per year or 800% a year. This 800%, I believe, is the theoretical limit for EOD NDX trading in Profunds. If I am correct about 20% a month as a theoretical limit, then 15-20% a month may be achievable in practice.

George,

Well it looks like a trend change by my definition although not a very convincing one.Still I have to follow it and buy at the next dip since we are now in overbought condition.Your thoughts.

Regards

Marc

George,

Have you archived your June and July trades on iHub?

sounds good. my new personal target - i'm not there yet, lol - is to capture 30% of the daily range...every day.