Coin tossing and random entries

Hi GeeSeven

So as not to further hijack the thread in the link http://www.forexfactory.com/showthre...38#post1492238, this is in response to your post here:

I respect many of Van Tharp’s views. I’ve read Trade Your Way to Financial Freedom, subscribe to his free e-mailed newsletters, and have watched a couple of his videos. However, just because he’s written a couple of best-sellers, and runs seminars across the US, doesn’t make him infallible, especially if people choose to quote him out of context.

On the surface, Van (and his mate Chuck LeBeau) apparently believe a profitable system can be built around random entries. But it should be pointed out that their views were publicized, in the context of trading stocks, during the strongly trending bull markets of the 1990s. When markets trend, accuracy in entry becomes less important.

Look at the chart below of the Dow Jones Index, especially the 1990s. The extreme case of letting profits run is “buy and hold”, and the chart shows why this policy paid investors off handsomely during this period. Even for swing and position traders, there are clear benefits, probability wise, in taking long positions and letting them ride. It should be mathematically self-evident that the more strongly prices trend, the greater the benefits of letting profit ride as long as possible.
It’s important also to define what is meant by random entries. Van, Chuck and others are talking about the timing of the entries, not the direction. If the coin toss being described in the link above had initiated short trades when the market is trending strongly upward (or vice versa) – which it would have done 50% of the time – losses are the likely result. I’m sure that Van would agree that the trend is indeed your friend, and a totally random coin toss entry system ignores this.

[Incidentally, in a recent video of Van’s that I viewed, he states that he doesn’t actually trade himself. If I understand him correctly, he is a psychologist and trading coach who allows a portion of his investments to be managed by his most successful students].

Forex brokers claim that one of the benefits of forex trading is that prices trend better than stocks’. Wanting to increase their client base provides a good motive for them to say this. If I may be permitted to generalize, nations’ economies tend to move in boom-bust cycles, which means that over time there is a certain amount of reversion to a mean. You can’t make a fortune using “buy and hold” over a lifetime, by “investing” in forex, like you can with blue chip stocks (I know a gentleman who died recently aged 94, leaving behind a $23m fortune, courtesy of a lifetime of “buy and hold” of a basket of NZX stocks, and simply exchanging the dividends for additional units). Moreover, with forex one is trading a currency pair, which amounts to the strength of one country’s economy relative to another. Not so with stocks. This suggests to me that different approaches are required.

Sorry if this sounds high-handed, but surely it’s unwise to take one quote from an author, and without providing anything in the way of an explanation, assume that it was intended to apply to all situations at all times. Like I said in my original post, if there was one simple rule that guaranteed trading success (e.g. random entry + let profits run), it could be rightfully packaged as the holy grail, and everybody would be using it. But trading is a zero (actually, negative) sum game, involving willing buyers and willing sellers, and these facts alone necessitate that different strategies are required for different situations.

I believe that, where there is no significant trend bias, entry is at least as important as exit. I respectfully submit the following posts for your consideration:
http://www.forexfactory.com/showthre...66#post1441266
http://www.forexfactory.com/showthre...77#post1452677

I'm very willing to eat humble pie. If anybody can show me a system based on coin-toss entry that delivers a higher profit factor, on balance across all market conditions, than my currently evolving method, I'd make the change without hesitation.

David

Coin toss entry:

Heads go short.

Tails do nothing.

This would be in accordance to THE BLACK SWAN. Stop loss prevents a runaway loss in the opposite direction and allows for a favorable BLACK SWAN to make us rich. Market conditions become inconsequential because price action will provide modest gains during retracements.

It would be interesting if you replace your present entry with coin toss and keep the same exit, money and risk management systems in place.

For some people ( me in my early days ) a Random entry is likely better than a planned entry as most people will try to pick tops / bottoms which will cost them most of the time, atleast with this you've got a 50% shot of shorting at a bottom for the move lower which almost always happens.

Certainly a easy entry system to teach or write a book about.


If you can't trade, write a book

My understanding is that a 'black swan' describes an event that's both unable to be predicted, and extremely rare. So the probability of the stoploss being hit before any 'back swan' event occurs would be extremely high. That's very similar to the purchase of a lottery ticket: an extremely high probability of a $5 loss, and an almost negligible probability of a life-changing gain.

The idea of constraining loss while letting profit run unencumbered can only be profitable to the extent that markets 'trend'. Such a strategy fails dismally in a sideways market. As with all trading strategies, it becomes a question of expectancy: do markets 'trend' far enough, often enough, to make such a strategy viable?

If I have an entry system that, on the basis of 1:1 RR trades, generates (for example) a 60% win rate, then it would make sense to me to use its signal as the basis of a short entry, rather than a 50/50 coin toss.

Here's something I learned from Ronald Raygun back when I first started snooping through the FF threads:

On your demo account: flip a coin and enter (your favorite pair) randomly with standard size - short=tails/long=heads. Do this x number of times in set time period t. I go min 20 entries in 30 minutes on the M1/10s charts - each trade must be 1 minute or less - but, obviously, I'm a short timeframe scalper. Goal is to maximize profit/minimize loss thru effective trade management.

I do this about once a week, usually around Monday Asian open. Good practice. I usually lose money.

The "Coin Flip" theory sounds good as an idea, but when placed live you will lose more than you win. Sure money management plays an important part but a bad luck of the draw can/and probably will keep you on the losing side for a very long time. If you think that increasing your risk by adding on LOTS, then you are trading nothing more than a Martingale system.

Has anyone every, picked a side, and stuck with it? As opposed to choosing random entries. Your chances goes up when you have a plan. For example, at trader that has been LONG biased for years will probably experience less losses and risk because he/she knows what to look for in the market (i.e. trend reversals and retracements) and knows which direction to take.

Traders that spend too much time trying to predict which way the market is heading will miss out. We all know, the market goes up and goes down, so why not just choose a side to stick with for the year and trade your plan. So, technically it shouldn't matter for short term spot forex traders to predict which way the market is heading. Understand that if you are a Long Term Investor, then yeah, direction matter. Intraday traders only need to pick a side ... I guarantee you will find an entry, all you have to do is wait. And if you are a trader that uses Fibs and indicators to "predict" the future, then I know you are experienced in waiting for a signal.

My signature pretty much explains it.

I'm a long time believer in the randomness of price movements. However, coin tossing your way into decision making is a non-option... even with my beliefs. Why bother with flipping quarters at your monitor, when you can go to Vegas and do the same thing while you have way more fun. Not to mention, I think that the odds at a roulette table would be slightly more favorable than your odds of overcoming broker spread.

Bottom line, if the coin toss is appealing for a trading strategy, try the following.

1. Slap an MA (you pick the period) on your favorite pair and TF.
2. Decide on flip cycle (say 1 flip every hour).
3. Heads = 1 lot, Tails = 2 lots. Or Heads = 50 TP/SL, Tails = 25 TP/SL.
4. Places orders on the appropriate side of the MA.

Bingo... your odds are now improved. And decision making is slightly more "logical" (for what it's worth).

Just a thought, although I'm fairly confident in saying that I don't think it makes a lot of sense to do anything along these lines.

MM

Do you believe a coin toss is truly random?

Of course not, they're all predetermined by the will of the Gods. Unfortunately, that does very little in the way of your P/L outcomes.

Okay, then that would mean that our success or failure as traders is also predetermined by heaven above.

Who said that praying for your trades is pointless?

We've found the grail!

coin toss results check that out......

That's getting a bit OT, we can use random.org or something if you really want to. a 51-49 isn't going to matter anyway really because we're most likely not going to be making millions of trades.

I like the quote in your signature.

Sorry, if I wasn't clear. The point is to protect against the negative black swan and to put yourself in a position to benefit from the positive black swan if a black swan occurs. In the meantime, you can be generating modest profits under any market conditions include sideways or range bound trading markets.

In the case of the coin toss system I propose, even if it only wins 50% of the time, the average win will me much larger than the average loss due to the cyclical nature of the markets.

Do you have a 60% win rate entry system?

It's possible that our concept of black swans differs. The only book of Taleb's that I've read is Fooled By Randomness, and it was a long time ago. I see a black swan as something so unlikely that it defies any kind of testing, and wouldn't be worth tampering with my setups to try to factor it in. If I miss an out-of-the-blue, once-in-a-decade move, then so be it. But of course no two people approach trading in exactly the same way.

I use at least three different entry setups that give me a significant positive bias. I can't say conclusively that they deliver a > 60% win rate, because very few of my exits eventuate at 1:1 RR, and of course trade expectancy is the product of both win rate and win size (RR). For example, one such setup involves trading pullbacks in a trend, placing multiple entries at what I see as being significant S/R levels. Many of these constituent entries end up generating trades with RR that's significantly > 1:1. So while the 'overall' win rate is probably < 60%, the entry points are nonetheless key, because they allow the tight stoploss placements that facilitate the potentially high RR. To what extent anybody would want to attribute its success to the entries or the exits is IMO purely academic; as long as it's overall profitable, I don't really care! However, making completely random entries would trade against the trend 50% of the time, ignore the timing benefits provided by S/R, rejection candles and relative currency strength, and consequently impair the effectiveness of this strategy.

If folk are profitable making coin-toss entries, then good for them; it must save time not having to look at price charts, and also eliminate any feeling of personal responsibility regarding their entry. All I can say is that I've not been able to make it work for me, or rather that I've found setups that suit my trading philosophy, and have to date proven themselves to be on-balance profitable. Each to his or her own.

No.......a coin flip is the equivalent of giving up. Just send your money to me. There is a ribbon of intelligence that runs throughout the markets. Traders are creatures of habit who do the same thing over and over again. Your job as a trader is to unravel that ribbon.

Yes, that's the way I understand it. I'd like to think that I'm finally on the way toward unraveling the ribbon.

As I read somewhere "forex is an activity where those who understand the markets eventually take all of the money from those who don't".

Is this a high(ish) probability setup for a long trade?
How well does it pass the 7-point plan?

1. Short term trend (olive colored channel in first screenshot) is UP. Tick.

2. Decent pullback in short term trend (below median line in olive channel). Tick.

3. Strong trend resumption candle? Large engulfing body. Tick.

(I've written an indicator that alerts me when points 1,2,3 occur simultaneously – sorry, it's not for sale. The rest of the process is discretionary).

4. Decent S/R below entry candle? Partial tick, but good enough. (Bands on chart explained here). Also has trendline support (manually drawn white lines). The objective is to place the SL in such a way that there are as many potential barriers between entry point and SL as possible.

5. Longer term trend/currency strength (see D1 chart in second screenshot): CHF (yellow line) strongly up. JPY (purple line) strongly down. Big tick. (That's why I chose CHFJPY over other pairs' setups that occurred at the same time.)

6. Has the trade got room to move (including the spread)? Probable targets T1, T2 based on prior S/R. (T2 is an obvious potential stall point if one zooms out to H4 chart). 15-20 pips on a M30 chart for this particular pair is good enough. Tick.

7. Check that no major news announcements are imminent. Tick.

Place buy limit order at the high of candle "A" which is obvious 'flipover' S/R point. Gives a low risk entry (SL and potential target areas shown). Sometimes when the setup candle closes near its high, I'll place a market order there, but I didn't feel that it was justified in this case (turned out that I missed an opportunity).

As it turned out, the order never got triggered.
But hindsight shows us that the target points were reached, while the stoploss was never threatened.

Was this entry setup better than 50/50? You be the judge.

Nice analysis, but I don't really know how to interpret the 2nd screenshot. Is it a graph showing the indexed currencies combined in one chart?

Each colored line represents a weighted average of the closing prices all of the broker-supplied pairs for the given currency. A rising trend means that a currency is gaining strength in the displayed timeframe, relative to its peers; a falling trend denotes increasing weakness. The displayed chart has its smoothing factor set to 4, which means that the values lag the price action slightly.

For detailed info, please refer to this thread. Post #1 gives operational instructions, and explains what all of the parameter settings do. Post #7 explains the calculations behind the colored lines. Post #20 attempts to outline some possible uses for the indicator, in terms of a trading methodology, although you'll note that I use it differently in the example above. There is further miscellaneous information scattered throughout the thread.

If, after reading the material, you have more questions, please post again, and I'll do my best to answer them.