Flip The Coin System

Too busy screwing around in F.F

Hehe...

I must be missing something here. To me it seems obvious that if "prices" truly moved randomly (as simulated by a coin toss), the probability of a "trend" ending at any point is equal to its ending at any other point. Or, put another way, there is a 50/50 chance that a "trend" will continue, or end, at any point. Hence there is no benefit in letting profits run, because expectancy (if we ignore costs) must necessarily be zero, regardless of whatever exit method is used.

The idea that the markets can be traded profitably (i.e. over an indefinite time period) is based on the assumption that there is a recurring aspect of non-randomness that can be identified, and trading rules (entries and/or exits) devised to exploit it. Examples of inefficiencies might include greater probbaility of reversal at support, resistance, Fibos, etc; or the idea that once a "trend" has begun then it is more likely to continue than reverse. These are of course brought about by collective psychology that can not be perfectly simulated by a random generator, however random their outcome might appear.

Conversely, any coin tossing simulation will produce what appear to be "trends". As an experiment, I once presented 3 real and 3 randomly generated charts to our local Technical Analysis interest group. Not one of the 20 or so attendees were able to score 6 out of 6 in terms of guessing which were real and which were "fake". I can post them here if anybody’s interested.

In general terms, if letting profits run enhances bottom line, then that is evidence that markets "trend" more often than not. The greater this trend bias, the more important it becomes to let profits run as long as possible. Conversely, if prices are ranging, then taking profit early, at or near an appropriate area of S/R (where probability of reversal is higher), becomes the optimum exit method. Explained more fully in points 1 & 2 in my post here: http://www.forexfactory.com/showpost.php?p=1452677

Re the likes of Tom Basso and Van Tharp conducting such an experiment, see my posts here:
http://www.forexfactory.com/showthread.php?p=208866
http://www.forexfactory.com/showthread.php?p=1492382

If anybody is interested, the random entry coin toss has already been discussed here:
http://www.forexfactory.com/showthread.php?p=1490989

Bullrock, I wish you good luck with your experiment, and I await the results with much interest. And thanks for posting here rather than the Trading Room, so that newbies like myself can respond.

David

Phil, if your friend is winning overall with a 40.1% batting average, then his ave win size must be significantly greater than his ave loss size, i.e. R-value > 1. IMHO this can be arbitrarily attributed to either a "let profits run" approach to exit (if he is riding trends), or his entry system's ability to pick setups that ultimately lead to decent moves..... probably both. Hence the entry is probably contributing to his success; if it wasn't, then he should be able to achieve the same return with random entry (which brings us back to the original topic). If his is a trend following method, then trading with the trend is not really a truly random entry.

I agree that conservative position sizing preserves capital (for obvious reasons), but would suggest that bet size has no relation to expectancy. The latter is determined solely by the efficacy of the entry and exit (which I assume is what is meant by "trade management") rules.

David

Yo.. BR.. hows the progress of ur strategy??
Or u just simply abandoned this thread already?? :

maybe someone could do this historically... pick random dates and time of day and see how it would have turned out over i dunno... 100 trades?

Instead of waiting for this to be done in real time.

Only limit is your intraday historical data.

Really interesting subject though. I believe that MM is the most important aspect of trading.

Hi. I am so f*** busy these days. I am gonna to launch this anyway soon. I let you know when. Probably in January.

I guess I should have read this page first!

This idea is interesting. Has anyone completed any forward testing? I would love to hear the results.

I think the result of a random entry + a trailing-stop MM system will just give you an indication of whether the market was trending or not.

Imagine two scenarios.

Scenario 1 - Bull Market
Tails - goes short - trailing stop takes you out
Heads - go long - you ride the market for profit
You make big money 50% of the time and lose small money 50% of the time.
Net = You Win

Scenario 2 - Range-Bound Market
Several different possibilities, but it's highly likely that the stop loss will hurt you. In many range-bound markets you want to do the opposite of using a trailing stop: taking profits while doubling up (Martingale system) on losses appears to be the best bet.
Net = You Lose (if you are using a trailing stop)

So my guess is, if you use a random entry + a trailing MM and tell me your Net result, I can roughly tell you if you have a trending market or a range-bound system. I think conversely, you can "probe the market" for being Range-Bound with a Martingale system.

Also, we are assuming here that the time-series is stationary (ie statistics like averages and volatility are constant). In normal, non-stationary price series, the results may be even less certain as sizing of trailing stops is very sensitive to the volatility of the underlying series.

Another thing to note: a random entry seems really impressive, but when you use some form of MM (either trending-style trailing stop or range-bound Martingales), there is "leakage" of market prices into your system.

So although your entry is free from market-information, your exit (based on mark-to-market), certainly isn't...

Rather than flip a coin, choose long or short based on a fixed line in the sand like a daily pivot line ( high + low + close ) / 3.

Keep SL at 10 pips. Let a winning trade run. We know the H1 average range exceeds 30 pips. The D1 average range exceeds 200. One win can wipe out 2 or more losses. Let's be reasonable, no one trades every hour of every day. The winning gamblers know when to get up from the table and cash in their chips.

Still Busy Mate?

Sorry but this doesn’t work
Markets are indeed random. But no MM can coorect that.
You have 50/50 chance of going to SL at 100pips or TP at 10 pips. But your lotsize will be smaller at TP100 since you areday closed part of it at TP50.
So no. Won’t work.

Do you still trade the news using your straddle method from 2016?