|
PureBytes Links
Trading Reference Links
|
clarification:
for trade series analysis ... probability (frequency), for any magnitude, is derived from the empirical cum freq dist and then the prob can be used to determine the chance of getting a run of wins or losses (drawdown+- == prob of a loss ^ #trades * possible geomean of trades).
Empirical == measured or observed values rather than theoretically calculated from an assumed dist (parametric).
There you go ... the forum now has the whole box and dice ... all of the evaluation theory that any trader could ever need.
All we need to do now is develop the working models .. privately or publically.
--- In amibroker@xxxxxxxxxxxxxxx, "brian_z111" <brian_z111@xxx> wrote:
>
> I don't think the discussion is slightly OT ... I think it is right on the money.
>
> Around the time that I first joined the forum someone told me off for discussion stats and/or trading psych ... they said that the forum was for dicussing and helping newcomers with code ... that is what the forum meant to them but to me it is also a place where we can discuss trading issues with real traders ... people who live and breath the subject .... you can't buy this anywhere.
>
> Forum disucussions often exceed what is available in the run of the mill trading books, especially in detail and scope.
>
> This is really a discussion about system design and evaluation/money management.... it can help people to maker better private applications, within AB, possibly influence Tomasz with design and maybe end up with some public applications (plugins or code etc).
>
> We can't produce good designs if we don't sort out the design philosophy in advance.
>
> <snip> At an average of $60/hr x 8 hrs/day (or, $0.25 x 1920 plays per 8-hr-day), the ('average') 'income' generated of $480/day 'pocket-change' is not too shabby.<snip>
>
> Yoy logicos ... honestly ;-)
> I figured out that it averaged around $40+ an hour in 2 secs with my calculator.
>
> Re equalizing:
>
> You and Tom are both talking in $ and reporting expectancy in $.
>
> I recommend you think about equalizing and perhaps adopt the habit of reporting in % e.g. if a stock is trading at $10 and you win $1 then you have a 10% retun ... if a stock is trading at $100 and you win $1 then you have a 1% retun ... there is a big difference.
> Also, using % for reporting allows us to standardize our systems so that we can compare, say, an option trade to a futures trade to cash or stocks.
>
> That is why I went to all of the trouble to remodel ProfitFactor and turn it into PowerFactor = PoF (the equalized, standardized, reinvestment version of unequalized constant contract PF)
>
>
> Re options:
>
> Someone asked recently about backtesting options in AB and getting the data in etc.
>
> I was too busy to comment but I thought it was all made unnecessarily difficult.
>
> Given that there are sophisticated option plays that involve mispriced options and arbitrage opportunities etc but if I am a 'game designer' then the following simplistic model seems to have some use:
>
>
> - an option can only win if the underlying moves so our strategy is derived from BT of the underlying first, to identify an edge (if we are selling options then of course we want to successfully predict non-volatility but in this case I am talking about trend trading by buying calls/puts)
>
> - to implement the strategy, using options, then the option premium can be considered as being analogous to taking out an interest only loan to 'buy' the shares we are controlling.
>
> If you look at the time to expiry and reference the 'premium' you can restate the premium as 'interest costs per bar' and so quite easily simulate the approximate outcomes of the trade (where the premium is the carrying cost)..... although an option is the equivalent of an interest only loan it is one with a fluctuating interest component but since it goes to zero value on expiry the mean nominal interest rate is predictable.
>
> Note that on this basis the interest rate you are paying is pretty high because it is actually an insurance premium against incurring the total possible loss (the market in your instrument goes to zero value ... assuming you are long).
>
>
> I think that implied volatility could tell us everything we want to know about the 'cost' of this 'insurance'.
>
> So, if we can get implied vol data (that we can rely on) into AB we can reasonably test strategies that use options as a vehicle for controlled risk/reward relatively easy (I retain a reasonable knowledge of synthetic strategies from my option trading days and I feel there is a good deal to learn about system design from understanding synthetic option strategies).
>
> On the other hand ... if people want to do sophisticated BT of option strategies (the Greeks etc) I believe it can be done in AB as long as we can get implied volatility data I think I am correct in saying that all of the factors can be backcalculated from the price and the implied).
>
> The difficulty with opt analysis is not in the math but in the data itself ... lots of simultaneous expiring contracts with the symbols being recycled at a later date (worse than futures?).
>
> Reasonable approximations might be possible if we could get data, like a continuous contract, or spot cash for futures, for the nearest price option implied volatility ... don't know if it exists and would have to think a bit more about the implications.
>
>
> Re your observation that the returns from trading exceed the returns from Blackjack, or Video Poker ... using the analogy that trading is a game of chance:
>
> - the reward in trading is far greater than any casino game but so are the risks
>
> - casino games never experience Acts of God or not to the same extent anyway ... I suppose if the casino throws you out for card counting at Blackjack it is a little similar to when the markets close (throw us out of the house), during times of turmoil, while the insiders sort it all out (of course unlike the casino owners they are impartial and never bias their decisions to benefit themselves or their friends)
>
> - they are very predictable while the markets have a degree of unpredictability about them .... the odds in casino games are predictable and the same everywhere ... in market games they are unpredictable and depend on the design of the game and the stability of the underlying... the distributions of the returns of casino games are by the book whereas in the market they have fat tails and often don't conform to any typical distribution
>
> - casino games are mechanical while the markets are driven by the behaviour of participants.
>
>
>
> Here is my 'off the top of me head' running total on the main classes of market risks:
>
> - design risk (our games aren't what we think they are ... only live trading can test the reliability of the model)
>
> - markets change, or are non- stationary (same model == different game)
>
> - variance, which is dualistic i.e. variance of the mean (which is best described by the GeoMean for trading with reinvestment) == Reward
>
> or if we extrapolate this, variance of the Wins/Losses ... this has two components = magnitude and frequency .... empirically magnitude is the best win/worst loss or other variations on the theme ... empirically frequency is probability, derived for the cum frequency distribution, which plays out as runs e.g. losing streaks (or drawdown) == Risk
>
> - Acts Of God (market turmoil closes the markets for 2-3 days)
>
> - administrative risk (failed admin systems/tools etc)
>
> - behavioural risk (ours and others e.g. employees at our brokers)
>
>
> --- In amibroker@xxxxxxxxxxxxxxx, "Buzz M. Ross" <buzzmr@> wrote:
> >
> > Tom,
> >
> > I found your commentary about the gambling rather intriguing. I'm inherently an 'applied mathematician', so the numbers you shared triggered some questions:
> >
> > Since your coin-turnover is about $4k - $18k per hour, let's take a 'high' average of about $12,000/hour (for ease of numbers), or about $200/minute turnover. My experience in video poker when in Las Vegas has been about 4 plays per minute (i.e., about 15 seconds per play to place bet, evaluate displayed initial hand, select replacement cards, and view outcome), give or take. If your experience is somewhat similar (though I suspect you take a bit less time per hand), this would translate to about $50 in coin per play, and if I understand the machines correctly, you can 'deposit' (bet) up to 5 coins per play. If this is correct, then it appears that the machines you use take $10 coins. I don't know of $10 coins, but I assume that there exist machines that allow '$10 per coin' bets and you're using the machine-stored 'cummulative-account' currency to make bets for each play. Has my analysis been 'reasonably' correct so far?
> >
> > If you play, say 3 machines at a time (sitting in front of the 'center' machine and having access to the left and right adjacent machines), then the analyis, of course would change a bit, but you still have constraints on the minimum time required (on average) to play a hand from bet to outcome.
> >
> > If my numerical analysis above is 'on target', then from a statistical standpoint, I would conclude the following:
> >
> > 1) Your RATE of transactional 'action' is, based upon a maximal $50/play bet, from 80 to 440 plays per hour.
> >
> > 2) The 'edge' of 0.5% would then evaluate to 0.005 x $50 = $0.25 (25 cents) per play of 'Expectancy', or about 4 x .25 = $1/minute = $60/hour 'AVERAGE', give or take, to sit at a Video Poker machine.
> >
> > 3) With an initial 'bankroll' of, say $6,000 (for numerical convenience), it would take about 100 hours (from #2 above) to generate $6k in 'gains' to Double the inital bankroll, or about 12-13 days of 8 hours/day in front of the machine(s). Assuming the 4 plays/minute x 60 min/hr x 8 hr/day x 12.5 days/double_$6k = 24000 plays! Of course, with an 'expectancy' of 25 cents/play, it takes 24000 plays to generate $6000 (to double, in this case, the initial bankroll).
> >
> > 4) At an average of $60/hr x 8 hrs/day (or, $0.25 x 1920 plays per 8-hr-day), the ('average') 'income' generated of $480/day 'pocket-change' is not too shabby.
> >
> > Just for fun, let's assume a stock option play as follows:
> >
> > A stock is trading at, say $50/share and is in an 'established' uptrend. Average true range (ATR) over the past 20 days is 'roughly' $1.00/day. An 'at-the-money' 'call' option with a strike price of 50 and expiration about 3 to 4 months 'out' (so that 'time decay' is small) is currently trading at a premium of, say, $5.00/share, and the implied-volatility is relatively 'moderate' and 'stable'. The 'delta', due to the 'at-the'money' combination of stock price and strike price, is about 50% or 0.50, so that for each dollar change in stock price, the option premium changes in the same direction about 0.5 x $1 = $0.50. Remember, the ATR PER DAY of the stock price recent history is about $1, so it might be expected that the option price will change about $.50/day in concert with stock price fluctuation. If 10 contracts of this option is purchased 'long', which is a 'position size' of 1000 shares, then it is possible for the stock to move 'favorably' (due to 'uptrend') about $1 in a day, and hence, for the option premium to gain 1000 x $.50 = $500 for that day. The 'worst-case' risk of loss, of course, is the 1000 shares x $5.00 acquisition premium = $5000 ('similar' to the 'typical' bankroll for Video Poker), but unless an 'unusual' event-shock day occurs for the stock, it is very unlikely that the full $5k commitment to the option would be lost.... a reasonable 'loss limit' decision would have been instituted BEFORE the trade was placed. So, for comparison purposes:
> >
> > 5) How much time would it ROUTINELY take to make this kind of evaluation and decision, and place entry and exit order(s) for this ONE transaction (i.e., 'play'),
> >
> > 6) How often can this kind of transaction be REASONABLY made over a 12 to 13 day period? Perhaps 1 to 3 of these 'plays' can be made per day, so there would be about 10 to 35 'plays' per 13-day interval. Compare this to the 24000 video-poker plays required to generate ROUGHLY the same 'revenue' (assuming the 'positive expectancies' can be made 'equivalent').
> >
> > 7) Imagine what outcome over time would accrue if, say 1/10.th the number of decisional 'plays' were made using the option scenario (2400 x $???, COMPOUNDED!!) versus the 24000 'robot-like' (for 'correct' play) decisional plays made for the video poker endeavor. (This does NOT, of course, take into account the emotional 'joy' of playing machine poker for 'hours on end' versus the emotional 'stress' of commiting 'capital-at-risk' to the market 'mechanisms'.)
> >
> > This is just some of my 'neural cogitation' at play!! (;->)
> >
> > However, Tom, I truly would be interested in understanding if my analyis of the numbers you presented about your experience in Video Poker is reasonable or if I'm 'missing' something crucial to that analysis.
> >
> > Thanks for your sharing and clarification(s)....MUCH appreciated!!
> >
> > Buzz
> >
> >
> >
> > --- In amibroker@xxxxxxxxxxxxxxx, "professor77747" <professor@> wrote:
> > >
> > > Brian,
> > >
> > > First, about my gambling, I only play video poker and only games which a positive return if play correctly. Most of the time my edge is only a little over.5%. Like today, I played at .22% on the game, but the will mail be another.33%.
> > >
> > > Since I average between $4000 to $18000 coin in per hour, it is a nice profit. However, I also get comps for rooms, restaurants, shows, cruises, and other things. It is really a good deal.
> > >
> > > I spent years playing with different ideas for gambling, but the only beatable games in a casino are blackjack and video poker. Some people are good at sports betting, but I am not one of them.
> > >
> > > About my formula, I use the system downdraw as a guide. I have had 2 maximum loses in a row a few times, but I would over about 5 in a row to lose the maximum system downdraw with my stops.
> > >
> > > I do worry about the system going down or losing my connection, so I check it at my trading times most of the time. I can tell if it may trade to I will check it. I have an email sent to my cell phone when it trades and then I check it out to make sure that it is correct.
> > >
> > > I have a laptop with me at all times so I can connect to the internet and then connect to my desktop to check if it is running correctly.
> > >
> > > I don't have much trouble except that I made a slight change in the formula and now it doesn't place my second trade, so I quickly login when I get an email to place the trade. I hope to have that problem solved soon. It was working fine until I made the change a few weeks ago.
> > >
> > > Thanks for all your comments. I will have to do some reading.
> > > Tom
> > >
> > >
> > >
> > >
> > >
> > > --- In amibroker@xxxxxxxxxxxxxxx, "brian_z111" <brian_z111@> wrote:
> > > >
> > > > This is cool ... we have fortuitously dropped onto a wavelength.
> > > >
> > > > > I am a gambler. I live in Las Vegas and gamble almost everyday. I >have been gambling for almost 13 years and have made a very nice >profit every year.
> > > >
> > > > Thanks for letting me know ... I don't need to worry about you know because your eyes are wide open and had to consider the possibilty that you were a kid with your first 50K or maybe a young married with three kids and only 50K to your name (it happens).
> > > >
> > > > There is a funny side to my gambling experience ... I had only read enough to know that the Martingale was doomed to fail .. the house limit kept me to 8 bets (doubled) .... $5 min ... I would wait for 3-4 in a row (can't remember which) and then start to double ... after loss one I was playing for breakeven or bust.
> > > >
> > > > The house should have got me several times, based on my game plan, but it just didn't come ... I was amused at how long I got away with this dumb play.
> > > >
> > > > I was on the low rollers table and I became a minor celebrity amongst the junior staff ... I won that ofen they thought I had some big method to win .... they must have thought I was memorizing 1000 spins and calculating the imbalance in the wheel or something crazy like that ... I could see them talking and looking from time to time.
> > > >
> > > > It is amazing how we make up fantastic explanations in the absence of simple truth.
> > > >
> > > >
> > > > <snip> When I gamble, I have a mathimatical edge just like the casinos have. However, when I invest, I don't have that edge so I am not as confident during a losing period. In gambling, I have had losing months of over $25000, but I always come back because of the mathimatical edge.<snip>
> > > >
> > > > I am an intuitive/action man .. my strength in trading is coming up with new ideas and analyzing what is behind the strategie etc ... any inconsistency in the theory is brutaly exposed (mainly done in private).
> > > >
> > > >
> > > > So, after a while of thinking about trading I boiled it down to the fact that a system was like a machine (albeit an algorithmic one) that produced numbers that we bet on (the numbers come out in series and recorded as %)..... the machine runs over the data (stockmarket returns) to produce the numnbers.
> > > >
> > > > The English are quite open about this analogy because their gambling houses moved onto the markets and said,"Yes, we will take bets on the market moves" i.e. spread betting (which incidentally, if it is operated well, is actually a more efficient way to trade than 'buying' the real shares via a broker/clearing house == very inefficient because there are too many middle men).
> > > >
> > > > Thereafter I worked away at gambling theory (a little bit), learning how to identify and measure the edge in trade ... nothing else really counts that much ... after that I moved on to MoneyManagement (how to optimize the edge).
> > > >
> > > > I am surprized you aren't aware of the precedents:
> > > >
> > > > - it wasn't Osborne ... I think it was Eduard Thorpe who influenced Kelly (coin toss maths) ... Thorpes paper is on the net bit there is no need to read it or Kelly.
> > > >
> > > > I recommend that traders go straight to Vince and optimalF etc if they want to cut to the chase when it comes to the theory of trading the edge.
> > > >
> > > > The amazing thing I found was that the current definitions of the edge and the methods to quantify aren't spot on ... probably they are too orientated to financial theory, which has different investment objective to traders ... also they operate in a different timeframe .. the short term timeframes that traders end up using (for good reason) tend to demand that we focus on what our edge is and how to quantify it ... we have less margin for error (because commissions erode more of our edge).
> > > >
> > > > Basically I had to design my own.
> > > >
> > > > CoreMetrics, BiSim and asymmetricalPayOff are examples of my efforts ... so far I believe the ideas are good, I have provided some insight into the edge etc but I am not certain about the mathematical integrity (not being a mathematician) ... I am not losing any sleep over the later because if the integrity of my current metrics fails I will just fix them up or make some new ones.
> > > >
> > > > <snip>However, I was really flying by the seat of my pants. Sometimes I would diviate from my idea because of my emotions. I then lost $22000 on 2 trades and had less than $10000 in my account. I decided to get a formula and auto trade it to eliminate any decisions and emotions. It doesn't work. I mean I still have the emotional stress, <snip>
> > > >
> > > > Logical primaries (people whose favoured psychological mode is thinking) tend to think that the Psychology of Trading is over-rated (that is saying it politely) and that algorithmic trading automatically eliminates any subjective elements from the trading room ... wrong! People use computers to trade and not the other way around (don't they know that Arnie triumphed over the machines).... we are all fallible and psychologically complex.
> > > >
> > > > <snip>I didn't use stops then.
> > > > but since I have refined my program and added stops, it seems to work fine<snip>
> > > >
> > > > I can state with confidence that the edge is best defined by the GeometricMean ... as long as you are reinvesting (going all up) .... if you want to learn a bit more start with Ralph Vinces first book ... notice that in his basic model the reward is the GM and risk is quantified by the largest loss (historically or empirically ... later he develops other models and investigates other ways to measure the risk.
> > > >
> > > > In short ....
> > > >
> > > > .... the mean is an asymptote ... the more we trade, the more representative our mean record is of the real edge.
> > > >
> > > >
> > > > ... unfortunately variance goes the other way ... the biggest loss gets bigger and the biggest losing drawdown gets bigger .... variance is an asymptotically inverse.
> > > >
> > > >
> > > > AND ... it gets worse ... the roulette wheel and the blackjack game are mechanical ... the odds are the same always in every house on the planet (subject to the number of cards in the deck etc).
> > > >
> > > > The stockmarkets are not stationary i.e. not mechanical ... the odds move around a little or perhaps a lot.
> > > >
> > > > Luckily there are ways to turn the stockmarket into a game of TwoUp .... mechanically limit the losses (this changes the game somewhat) ... in the US this can be done by using options, using stops and maybe others I haven't found yet ... outside the US we can add guaranteed stop losses available through spread betters (CFDs') ... we should look at the security of our stops though ... if there is an act of the market Gods will my stop get blown away (maybe I am left hanging for a day or two while it is all sorted out?).
> > > >
> > > >
> > > > Note that my BinomialSimulation model is at worst very educational .. it models the market as if it is a coin toss with different PayOffs on the head and the tail.
> > > >
> > > > > I am just trying to learn how to manage my risk better to improve >my formula.
> > > >
> > > > I am guessing a bit but it seems to only need tweaking ... better check if there is any way your stops can fail (broker acts etc .. computer failure if AT) all of this equals risks maybe worse than some other obvious ones.
> > > >
> > > > Also, what will happen if you get the biggest loss two in a row or what will happen (to your account that is) if you get the mean loss 2,3,4 etc in a row ... assuming that losses tend to converge on the mean since if they are randomly selected.
> > > >
> > > >
> > > > Another risk, not found in card games, is that one big loss may have an influence towards causing a second loss in a row ..... above and beyond the normal odds.
> > > >
> > > > There is a lot in it and it takes a lot to boil it all down to a few pages ... Vince is a very good start.
> > > >
> > > > --- In amibroker@xxxxxxxxxxxxxxx, "professor77747" <professor@> wrote:
> > > > >
> > > > > Brian,
> > > > >
> > > > > I am a gambler. I live in Las Vegas and gamble almost everyday. I have been gambling for almost 13 years and have made a very nice profit every year.
> > > > >
> > > > > When I gamble, I have a mathimatical edge just like the casinos have. However, when I invest, I don't have that edge so I am not as confident during a losing period. In gambling, I have had losing months of over $25000, but I always come back because of the mathimatical edge.
> > > > >
> > > > > In investing, I don't know if I will come back or not. I am fortunate that I have made enough profit that I am working on my profits.
> > > > >
> > > > > When I first started investing, I studied the charts and had an idea for investing. It was just slightly profitable making an annual 2% profit. Than about 5 years ago, I started looking at gold futures and my profit was much higher. I ran $10000 up to $30000 in my first year of investing in gold futures which was about 3 years ago. However, I was really flying by the seat of my pants. Sometimes I would diviate from my idea because of my emotions. I then lost $22000 on 2 trades and had less than $10000 in my account. I didn't use stops then.
> > > > >
> > > > > I decided to get a formula and auto trade it to eliminate any decisions and emotions. It doesn't work. I mean I still have the emotional stress, but since I have refined my program and added stops, it seems to work fine.
> > > > >
> > > > > I am just trying to learn how to manage my risk better to improve my formula.
> > > > >
> > > > > Thanks,
> > > > > Tom
> > > > >
> > > > > --- In amibroker@xxxxxxxxxxxxxxx, "brian_z111" <brian_z111@> wrote:
> > > > > >
> > > > > > Tom,
> > > > > >
> > > > > > Thankyou that helps.
> > > > > >
> > > > > > Going to the forum is like going to the Doctors ... the more you tell us about the system the better chance that we can help.
> > > > > >
> > > > > > At first I wasn't sure if you were genuine but I remember you know and you are a very genuine, open and modest person.
> > > > > >
> > > > > > We talked a little once before .... Razbarry helped you with your AT algorithms etc.
> > > > > >
> > > > > > A couple of years ago Louis asked about what the realistic expectations for a trader should be and when I said "25% a year is the min, as the becnhmark", some people disagreed and said I was overstating the case.
> > > > > >
> > > > > > I recall that you came to my rescue by saying you were doing around 50%PA ... I think that was your first year of trading, or maybe your first year of AT.
> > > > > >
> > > > > > IMO this is first and foremost a case study in Trading Psychology more so than technical issues ... although one reflects on the other.
> > > > > >
> > > > > > So, the general tone of the advice you have received so far is correct....... you are the only expert, with regards to yourself.
> > > > > >
> > > > > > Perhaps we can help with a second opinion.
> > > > > >
> > > > > > However ... it is only an opinion:
> > > > > >
> > > > > > - I, and the forum, can only read so much from a few short posts,
> > > > > > - it is too much responsibility to make life changing decisions for others,
> > > > > > - I am reluctant to say too much, since you are going so well ... I say anything that detractrs from your success so far.
> > > > > >
> > > > > > Please keep in mind that the issues you are canvassing here are bigger than Ben Hur ... your case history could make an interesting trading book all by itself.
> > > > > >
> > > > > > We can't discuss every single issue in detail, and you won't be able to process it all in a short space of time, (processing is done in different ways by different people) ..... we can make a start ... you will have to break it down into bite size pieces, digest a bit at a time and maybe come back to the forum for some more at later stages.
> > > > > >
> > > > > >
> > > > > > ******************************************************************
> > > > > >
> > > > > > Sport is a metaphor for life ... for me (my wife doesn't think so ... she is a musician and so for her music is the key to life).
> > > > > >
> > > > > > Trading is a high performance event.
> > > > > >
> > > > > > We are all unique ... our trading styles are as individual as our fingerprints (recently Dennis and I swapped a couple of trading insights in the forum ... in many ways he has a similar approach to me == 1/10,000 forum members, but in other ways we can't agree at all).
> > > > > >
> > > > > > Some things we all have in common though.
> > > > > >
> > > > > > High performance demands aptitude; honed by theory, practise and performance. in a cycle of continuous improvement.
> > > > > >
> > > > > > You seem to have the knack for it (aptitude) ... in a very short space of time you have wandered into the winners circle, almost without any self awareness.
> > > > > >
> > > > > > I have had to work my way into the winners circle, self consciously, by philosophising and studying.... I feel a bit like a kid at College ... even though I got good marks last year and have done some good original research I look up and see a mountain of text books still on the table in front of me ... miles more still to learn.
> > > > > >
> > > > > > If we don't have the aptitude in the first place we will never make it, no matter what.
> > > > > >
> > > > > > If we have an abundance of aptitude we can do quite well without as much theory and practise as others.
> > > > > >
> > > > > > Also, we all have a weak link, and so we can go so far without making the effort to strengthen our weak link (usually we will avoid our least favoured aspect when really we should make it our priority to strengthen it).
> > > > > >
> > > > > > I think that at the moment you are a little like a College sportsperson who easily made it to the big league without even trying ... while 1000's of other hopefuls, who busted their arse, are still back there somewhere:
> > > > > >
> > > > > > - I hear that this it is difficult to make the transition from 'chasing the dream' to 'living the dream' .... it is different once you get there (the challenges of staying there are different to the challenges of getting there).
> > > > > >
> > > > > > - because you are a little unselfconscious about it all, and moved from the amateur ranks to the pros, in such a short space of time, you haven't 'paid your dues', the way many of us have and so know you have to live with that.
> > > > > >
> > > > > > Perhaps you don't realize how good you are and/or will have to do some catchup on 'making the hard yards'.
> > > > > >
> > > > > > Note that the things that a professional golfer is concerned about are a world apart from the things that amateur golfers think about ... a pro may study every little aspect of his game (mental and physical) ... redesigning his/her game and repetitively working on the detail to improve.
> > > > > >
> > > > > > As I said ... this is only an opinion ... one opinion ... take it all with a grain of salt.
> > > > > >
> > > > > > You might just be one of those traders who remains a little un-selfconscious about it all.
> > > > > >
> > > > > >
> > > > > > A few specifics and observations (not in order of priority):
> > > > > >
> > > > > > - I believe you are quite smart and technically competent but you don't blow your own trumpet
> > > > > >
> > > > > > - you know about the metrics ... you can do the metrics ... but you don't actually use any of them except for Profit% and DD ... perhaps there is a lesson there for you ... they seem to be your metrics of choice ... you are in good company because Howard is exceptionally well studied and knowledgeable on the Quants and CAR/MDD is his chosen metric (or one of them)..... if it ain't broke why fix it.
> > > > > >
> > > > > > - if you are really asking a question about metrics (after all of this time suddenly you want to dig into them?) then we (you and the forum) will need to go step by step ... you can't possibly absorb Quantsy knowledge and skills overnight).
> > > > > >
> > > > > > Study one metric, or evaluation idea, and ask questions about it and then move onto the next.
> > > > > >
> > > > > > - perhaps you are really asking something else ... you seem to be questioning your own tolerance for risk ... there is no objective metric to measure this ... it comes from inside and you are the only one who can see inside yourself.
> > > > > >
> > > > > >
> > > > > > A few reflections on 'tolerance for risk' to get you started:
> > > > > >
> > > > > > - for some people it is innate ... I am risk avoiding machine ... it is just in me .. I can instinctively taste, see and smell risk ... I see risk in places you guys can't even imagine.... you can't buy that skill.... just do your best with what you have.
> > > > > >
> > > > > > - there are formal risk assessement techniques and processes etc ... I worked in an industry where we ate risk assessments for breakfast ... formally recorded risk using written procedures.... some knowledge of formal techniques might help some .. this knowledge is in my subconsious and coupled with my innate feeling for risk it must help me a good deal.
> > > > > >
> > > > > > - I am not a gambler ... it is interesting that our money management strategies were honed by gamblers ... Osborne -> Kelly -> Vince et al and from betting on coin flips (we call that two up in Australia) to betting on a move on the price of Gold.
> > > > > >
> > > > > > - 20/30 years ago I went into business for the first time .. I was young and silly but somehow I realized that it involved risk and coping with risk .... this seemed to be something I had to test rather than intellecutalize.
> > > > > >
> > > > > > I hate Casinos ... noise, smoke, a fair % or the people are pretty rude, some big drinkers etc ... but I went quite often for a few months .. gave myself a kitty to start with ... the rule was to leave if I lost the kitty ... used a totally dumb methodological approach ... played roulette on the red and the black .. doubled the stake after a loss ... hadn't even read the textbooks about gambling and staking ... know a lot more about the odds now.
> > > > > >
> > > > > > I won week after week ... only small amounts ... got bored in the end ... went one evening and played like James Bond ... lost the lot in three hours (it wasn't a lot of money), walked out and never went back.
> > > > > >
> > > > > > Here is what I found:
> > > > > >
> > > > > > - I can stick to my game plan
> > > > > > - things don't get out of control if you have a plan when you go in
> > > > > > - you can't learn emotional control etc ... or be coached in it ... your temperament is your temperamement ... you either have it or you don't.
> > > > > > - when money is involved it seems like life and death .. normally nice people argue over a two dollar bet
> > > > > > - idiots around me and the noise are very off-putting when you are focused on the game and your strategy ... it is hard to block out the noise ... one has to make an effort
> > > > > > - some people are innately bad losers ... kick the table and fight with the staff
> > > > > > - every one is an expert ... everyone has their pet theory .. they love to talk gambling and especially give you some free coaching
> > > > > > - some girls liked the excitement and the money (I was approached a few times)
> > > > > > - newcomers distracted me from the game asking for tips etc ... I must have looked like I knew what I was doing ... I looked pretty serious writing down the trade sequences etc
> > > > > >
> > > > > > - no matter how much money is a stake ... the blood pressure goes up .... whether or not you can keep the blood pressure down is a natural skill ... perhaps we can learn to manage it better if it is not our natural bent.
> > > > > >
> > > > > > As I said ... all skills we can't really learn ... I guess if we devised a list of simple rules, practised them and stuck to them it would help those who weren't born as 'cool as a cucumber'.
> > > > > >
> > > > > >
> > > > > > You might be asking yourself those type of questions, via the forum, rather than asking questions about metrics.
> > > > > >
> > > > > > Yu might just sale through this step almost unconsciously or you might have to think about some new stuff, to give you a point of reference from which to make your decisions.
> > > > > >
> > > > > > Another relevant factor:
> > > > > >
> > > > > > - no man/woman is an island
> > > > > > - our environment affects us ... at different stages of our lifes we have different duties to face up to .. and we are living out a different life story.
> > > > > >
> > > > > > I am 58, semi retired .. . my kids are all grown up and have left home ... my wife has her own career .. she doesn't get involved in the details of my trading ... I am a free man, in most ways (spiritually and philosophically etc) ... I have a life outside of trading ... I have goals in my life that are not based around trading ..... trading is a means to an end ... albeit a wonderful trading ground to develop my temperament and skills ... all of this interacts dynamically.
> > > > > >
> > > > > > Perhaps these are the questions you have to ask yourself:
> > > > > >
> > > > > > - you say you can't take those losses (if they happen) ... are you 100% sure about that?
> > > > > >
> > > > > > - is trading life and death OR more important than that (famous quote from a US sports coach)
> > > > > >
> > > > > > - if you do lose your stake will you really blow up or not (commit sucicide as some do)?
> > > > > >
> > > > > > - are you playing with money you can afford to lose?
> > > > > >
> > > > > > - do you have a good job ... will it really matter if you lose at trading (financially speaking) ... do you have a partner, kids,sick parents etc that are dependent on your income? ... if so how will they react if you blow up?
> > > > > >
> > > > > > - is making money your life goal or a means to an end?
> > > > > >
> > > > > > - is making money the primary reason you trade? or is the fun of the game the real reason you trade? ...if you are really playing the game for the games sake then you probably won't explode if you do the dough.
> > > > > >
> > > > > >
> > > > > > Finally - if doing some more work on your technical knowledge of metrics, evaluation and money management will help you make your decision then you have a lot of work in front of you.
> > > > > >
> > > > > > Note that in the past you got to where you are without it.
> > > > > >
> > > > > > If you want to learn more about MM etc then Howards book, the forum and other books/sites definitely help ... it has taken me years and I am still not an expert in the subject ... I can talk to anyone about it without embarressment though.
> > > > > >
> > > > > > As an aperitif ... a couple of points on metrics, risk and MM:
> > > > > >
> > > > > > - there is a theoretical (mathematical) stake for every system and account size (a lot of hours study required to learn this stuff) but there is a corresponding optimal emotional level of risk ... where is your emotional level... really ... not pretend level ... you can't fake it ... figure it out now before going any further.
> > > > > >
> > > > > > - your biggest loss to date is not the biggest loss you will ever experience
> > > > > >
> > > > > > - the aim of MM is to stay in the game ... we can't win if we have no money left to play with
> > > > > >
> > > > > > - the size of the account is very important ... the bigger our account the further we are away from 'isk of ruin' ... in trading $10k is just training fees ... if you lose that much consider it the cost of your training course .... $100K is the minimum stake required to play the game (same as the tables in the casino) ... if we have less than $100k we go to the low rollers tables (options etc ...easy on the margin trading) ... at least $200-300K is needed to even get into the high rollers room.
> > > > > >
> > > > > > - if you are ruined do you want to walk away with half of your stake or are you really happy to play hard, play to win and wala away with zero dollars if you are ruined
> > > > > >
> > > > > > - if you trade on margin are you really happy to walk away owing the broker money if you are ruined ... maybe have to hold a fire sale to pay your debts?
> > > > > >
> > > > > > - looking at your systems, in some cases, you have 0.6 or maybe 0.55 prob of a loss ... six in a row == 0.6^6 ... and you have very high ave losses (trading futures on margin?) ... so far you haven't experienced that kind of losing sequence?
> > > > > >
> > > > > > This is not emotional Q factor .. this is a rational probability (probabilities happen with more certainty than we think they will)
> > > > > >
> > > > > > Re the metrics:
> > > > > >
> > > > > > - if you want more help please ask ... one at a time ... the forum should be able to assist
> > > > > >
> > > > > > - most,or all of them, have had some thought put into them so they all tell us something ... whether you place a value on what they are telling you is another matter
> > > > > >
> > > > > > - I have made some psuedo tickers from random data (say 100 tickers) ... then I optimized a MACrossever system on those stocks ... naturally enough, while the population is break even (if you join all of the psuedo tickers together they add up to a sideways market without significant up and downs in between ... if I look at individual stocks some outperform (the individual samples have variance ... half are up and half are down ... how much up and down depends on the volatility I introduce to the process when I make the psuedo tickers) .... after I optimize if I sort by any metric in most cases the same tickers end up at the top i.e. a large number of AB's builtin metrics are all telling us the same, or similar things .... looking at the exceptions, then looking at the 'correlated' metrics would be one way to start to learn something about metrics ... why are some similar and others different ... I guess we could study the exceptional metrics and either accept or reject them, for our own use ... that would either narrow the field, byu getting rid of some, or turn up something we like.
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > > --- In amibroker@xxxxxxxxxxxxxxx, "professor77747" <professor@> wrote:
> > > > > > >
> > > > > > >
> > > > > > > Brian,
> > > > > > >
> > > > > > > Here are the reports for my 2 formulas. The profit doesn't seem much
> > > > > > > different, but this year, the profit shows a difference of over $5000.
> > > > > > > The first one is the one that I posted at Report 1
> > > > > > > <http://success101.biz/Backtest%20Report.htm> . The second is the less
> > > > > > > risky one at Report 2
> > > > > > > <http://finance.groups.yahoo.com/http://success101.biz/Backtest%20Report\
> > > > > > > -15.htm> .
> > > > > > >
> > > > > > > What metric do you use to decide that you (your systems) are profitable
> > > > > > > ... annual % return or what? I just look at the profit %.
> > > > > > >
> > > > > > > When you say your system is "risky" what exactly do you mean ... is this
> > > > > > > a hunch or do you use a metric to quantify risk? I look at the system
> > > > > > > downdraw. I started with 5000 over a year ago and ran it up to $30000,
> > > > > > > but I was lucky. I then took out $10000. I really needed $20000 to
> > > > > > > start. You will see that the system downdraw is over $11000. I trade
> > > > > > > gold futures and a contract requires about $6000 so I need at least
> > > > > > > $17000 to trade one contract.
> > > > > > >
> > > > > > > What do you mean when you say that both systems backtest nearly the
> > > > > > > same except for "price"? .... what is the same after a backtest? ...
> > > > > > > what price are you talking about? Since I trade using a 6 hour bar and I
> > > > > > > only trade at the close of the bar, the backtest results are the same as
> > > > > > > my trades except for a slight difference in price sometimes when I trade
> > > > > > > at different price than the backtest price. What I meant was that the
> > > > > > > system trades and the backtest trades are almost identical. I auto trade
> > > > > > > my formula and I use market orders so the trades are placed fast. The
> > > > > > > biggest difference between the actual trade price and the backtest trade
> > > > > > > price has been 20 cents. While this is a $20 difference in my cost, I
> > > > > > > still make a good profit on the trades. I could probably save by
> > > > > > > manually trading and placing limit orders, but I like to auto trade so I
> > > > > > > don't have to worry about watching the computer at my trading times
> > > > > > >
> > > > > > > The reason that I want to know about the risk is that I am going to
> > > > > > > start trading 2 contracts now that I have $40000 in my account. The
> > > > > > > difference in profit between the 2 formulas is over $10,000 in a 18
> > > > > > > month period, but most of that difference was this year. The system down
> > > > > > > draw was over $11,ooo on the more profitable formula and only about
> > > > > > > $8000 on the less profitable one.
> > > > > > >
> > > > > > > I must say that I was ready to quit when I lost over $11,000 so that I
> > > > > > > why I am worried about the risk. On 2 contracts, that would be $22000. I
> > > > > > > don't know if I can handle it.
> > > > > > >
> > > > > > > Thanks for your help,
> > > > > > >
> > > > > > > Tom
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > --- In amibroker@xxxxxxxxxxxxxxx, "brian_z111" <brian_z111@> wrote:
> > > > > > > >
> > > > > > > > Hello Tom,
> > > > > > > >
> > > > > > > > The definitions, for the metrics, are in Howard Bandy's QTS book.
> > > > > > > > I believe they are also in the help manual (not certain about that).
> > > > > > > >
> > > > > > > > That is a good place to start.
> > > > > > > >
> > > > > > > > In my experience we have to make the metrics our own, so to speak, by
> > > > > > > learning how they are calculated and then trying to understand what they
> > > > > > > are measuring (compared to each other).
> > > > > > > >
> > > > > > > > So, you are trading successfully for at least a year and don't
> > > > > > > understand the statistics ... I don't know if that is good or bad or
> > > > > > > something else but it doesn't appear to be a problem for you.
> > > > > > > >
> > > > > > > > Anyway, it is good that you want to keep learning and extending your
> > > > > > > knowledge, even though you are already successful.
> > > > > > > >
> > > > > > > >
> > > > > > > > I don't know if I can help but please clarify your question further to
> > > > > > > give me, or others, a fighting chance:
> > > > > > > >
> > > > > > > > Please post the report for each system over the same time range.
> > > > > > > >
> > > > > > > >
> > > > > > > > What metric do you use to decide that you (your systems) are
> > > > > > > profitable ... annual % return or what?
> > > > > > > >
> > > > > > > >
> > > > > > > > When you say your system is "risky" what exactly do you mean ... is
> > > > > > > this a hunch or do you use a metric to quantify risk?
> > > > > > > >
> > > > > > > > What do you mean when you say that both systems backtest nearly the
> > > > > > > same except for "price"? .... what is the same after a backtest? ...
> > > > > > > what price are you talking about?
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > > --- In amibroker@xxxxxxxxxxxxxxx, "professor77747" professor@ wrote:
> > > > > > > > >
> > > > > > > > >
> > > > > > > > > I have a very profitable formual that I have been autotrading for
> > > > > > > over a
> > > > > > > > > year. However, it is also risky. I have another formula that is not
> > > > > > > as
> > > > > > > > > profitable, but is also not as risky. My formula trade almost
> > > > > > > exactly as
> > > > > > > > > a backtest except for the price which varies by so little that it is
> > > > > > > not
> > > > > > > > > a factor.
> > > > > > > > >
> > > > > > > > > I don't understand any of the risk % factors in the top section and
> > > > > > > the
> > > > > > > > > factors below the drawdown figures in the bottom section.
> > > > > > > > >
> > > > > > > > > Here is a link to the statistics for last year which are very
> > > > > > > similar to
> > > > > > > > > this year except that there is more data. Statistics
> > > > > > > > > <http://success101.biz/Backtest%20Report.htm>
> > > > > > > > >
> > > > > > > > > Please help me understand these statistics. Thanks
> > > > > > > > >
> > > > > > > > > Tom
> > > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > >
> > > > >
> > > >
> > >
> >
>
------------------------------------
**** IMPORTANT PLEASE READ ****
This group is for the discussion between users only.
This is *NOT* technical support channel.
TO GET TECHNICAL SUPPORT send an e-mail directly to
SUPPORT {at} amibroker.com
TO SUBMIT SUGGESTIONS please use FEEDBACK CENTER at
http://www.amibroker.com/feedback/
(submissions sent via other channels won't be considered)
For NEW RELEASE ANNOUNCEMENTS and other news always check DEVLOG:
http://www.amibroker.com/devlog/
Yahoo! Groups Links
<*> To visit your group on the web, go to:
http://groups.yahoo.com/group/amibroker/
<*> Your email settings:
Individual Email | Traditional
<*> To change settings online go to:
http://groups.yahoo.com/group/amibroker/join
(Yahoo! ID required)
<*> To change settings via email:
mailto:amibroker-digest@xxxxxxxxxxxxxxx
mailto:amibroker-fullfeatured@xxxxxxxxxxxxxxx
<*> To unsubscribe from this group, send an email to:
amibroker-unsubscribe@xxxxxxxxxxxxxxx
<*> Your use of Yahoo! Groups is subject to:
http://docs.yahoo.com/info/terms/
|