Examples of Systems with an Edge

G'day All,

The aim of this thread is to discuss trading setups/situations which over time have a greater than 50% chance of being profitable with a R:R ratio of 1:1.
Sounds simple!

As trader who has really struggled at times and is really not profitable, I'm open to wisdom and experience from more profitable approaches to a trade.
I'm happy to help determine a setup's edge as I have a fair bit of data crunching experience.
Any setup or idea is worth discussing.
I'll get the ball rolling.
As you can see on the chart below, i had a successful long trade. I exited my long shortly into the candle after a bearish candle. I has assumed that based upon the previous price action represented by the candle that price had been going up and has failed and commenced coming down, so I exited a winning trade at the time happy with the pips I had made. However as you can see, shortly after I exited price rebounded very strongly and has continued to go up.

Based upon the information I had I chose to exit the trade (or I could have opened a new short trade based upon the price action of the previous candle).
Based on this setup, should I have expected price to fall? If so what % chance of the next trade being a profitable trade (based upon R:R 1:1).

I'm hopeful that I have explained the situation in my OP correctly for some great discussions and ideas.

Over to the FF community!!

Cheers
Timmy

Ordinarily I would have close the trade as well. But in a situation whereby am going long I would have let it run the next candle to be sure of my analysis by placing my stop loss in positive (some place half the previous candle).

Waiting for your opinion; FF Community!

I don't think such a thing exists.

That's going to cut the field down dramatically, I think?

The "net profitable expectation" systems I've seen, and use, tend not to have 1:1 R/R ratios, and of course there are reasons for that.

Just to clarify, you're interested in discussing only systems with a greater-than-50% win-rate?

That's not a complaint or a criticism at all, I hasten to mention - I'm just seeking clarification: it's an unusual criterion, certainly, given that most systems with an edge have a win-rate below 50%, for all the reasons explained by authors like Van K. Tharp (Trade Your Way to Financial Freedom), Tushar S. Chande (Beyond Technical Analysis) and Michael Harris (Profitability and Systematic Trading).

Great question and thanks for responding. My main objective is to review and explore "an edge". Doesn't have to be 50% expected win rate we can explore other options. Just thought by using 1:1 ratio was easier to manage the trade with set SL & TP.Open to suggestions and ideas

Thanks mate fair call you have a valid point. I waited a few minutes to see that price was beginning to fall on the new candle then closed trade. Your suggestion of setting stop loss while still in profit is valid.I guess in the context of this thread would we expect based on bear candle that price would be lower after next two candles? Would we expect this to happen more than 50%

A study was done where they took thousands of trades with a coin flip.
R:R of 1:1 and 1:2 was about break even.
R:R of 1:3 and 1:4 was nicely profitable.
(Results are somewhere on babypips.)

To improve on the coin flip(50%), put a 50 EMA on any daily chart.
Pick any time of the day every day (0:00 for example).
Buy if price is above MA and Sell if price is below MA.
Use R:R of 1:1 so on daily approx. 200 SL & 200 TP for most USD pairs for example.
No new positions until current position hits TP or SL to prevent a pile of positions during consolidation periods.

Should be better results than coin flip IMHO.

Well if you are going by pure price data, what common filters there are? (lot of indicators are built on these basic building blocks)

-time
-price change / stdev in some time interval (bollinger bands use stdev)
-volatility ( & oscillations/fluctuations amplitude and frequency)
-gap (price change with minimal fluctuation)
-patterns (moves with mathematical relations or technical analysis patterns)

But what is the premise in all this? If we are talking about backtesting an edge, the problem is that once a decent amount of real money is involved there is or may be hedging related to that trade. If the backtest spreads entry over weeks/months then maybe you avoid this problem. Otherwise you might need intraday measure of liquidity. You could use some of the above as filters in a model to estimate liquidity and restrict the backtest to trade only at liquid prices times. Then you could have a reasonable chance that your backtested model has some chance of working with bigger trades (I would still not go all-in in single trade!).

Much of the freely available technical chart trading advice may seem like non-sense but actually a lot of it is valid - the problem is the people explaining it don't really explain what the logic is behind the price action so I personally only started believing in some patterns once I understood what causes them. eg. Head & shoulders was visible just before the aug.24 drop. The patterns present certain market state but often some trigger is needed for the pattern to come true. eg. if Icahn comes and says something, that's probably such trigger but if the market state was not "ripe for it", even Icahn will have hard time moving the markets. This same logic applies to bad/good news - sometimes they work quite reliably in same way but that's conditioned on the market state at the time of the news.

If you read that carefully, I pretty much have away a possible 51+% 1:1 system. But GOOD LUCK TRYING TO DETECT THOSE PATTERNS WITH ALGOS! I've almost given up on it now that I know how complex they are and require WORK (charting tools) to even find in the charts! Now I just wait for the pattern and then take a shot! That's how most live traders seem to do as well. (the easier algo stuff that works tends to go into the HFT and commissions, spreads etc can get to you there)

(edit: Actually I've seen some quite impressive pattern finding stuff developed as automated chart drawing tools but incorporating such to backtests well, I haven't pursued it yet - and once such was publicly shared and everyone used same EA you could end up with liquidity issues and some "smart money" trading the other way and then any edge might become more random)

For this example perhaps a count back line stop loss would have helped you decide. I beleive there is a thread or 2 on the subject here at FF if you wish to checck it out.