Important – Why systems are profitable (or not)

Hi Dave,

I 100% agree with your statement below.
I have to admit that i tried hedging many moons ago in my rookie years but found that at the end of the day whether hedged or not it all came down to making a 'DECISION'.Hedging is not the edge its a way of managing positions.What gets you in/out of those hedged positions is your edge.
You don't make money because your hedged,you make money from reading the plays correctly imho.

I think many know that forex is a very different beast to other markets in its dynamics and so rather than hedging i now place wide stops and trade trends.

We are all different i guess and its only natural that we approach the markets in a different way,each to their own.To be honest,after trying hedging,i hated it.
I do one shot one kill.That's what i am comfortable with.
BTW,i just wanted to say that i enjoy reading your posts,keep it up.

Cheers
Steve

Hi David,

You've made some very valid points, I suppose alot of hedging can come down to feel, and what kind of market your facing that day/moment.

Hedging does give you time, nomatter how you look at it, yes you may take the same trades and yes you may make the exact same net result in profitable pips as me, but my cost of doing business maybe a few pips higher.

However, this is unlikely but possible, I'm not saying what you've explained couldn't happen, but you would have to be very exact on every extry and never suffer a fakeout and I personally believe that most people wouldnt be able to do this for any length of time. By hedging the trade in someways it takes away that pressure of been right first time all the time.

In my mind their is no doubt that I can be more relaxed in my entries, I think you'd agree theirs not much debate with that.

I can place a hegde anywhere in that channel and wait for the break, is that lazy ? uurrrmmm I guess, does it make much difference ? no not really and if I can make my life alittle easier then I will even if it does cost me a few extra pips.

The Hedging debate will always linger on, I don't use this method all the time as I've said, but in the right market it can and does make money. Its a more relaxed way of trading in my opinion as it does ( if nothing else ) give you more time to think and plan, unlike the normal way of taking a loss, I can and do wait for moments where the market seems more favourable to release my loser, so it does in my opinion offer me an edge all by itself.

So in answer to the questions, I do believe it offers an edge.

Its sounds good making the perfect entry day in day out, but its an unlikely event and so the fakeout element thats impossible to measure in my view does play a part in the math, how much ? its impossible to say.

I believe by hedging I can take away some of the pressure on the entry which I think is an advantage also.

And finally, I hope you'd all agree, the hedge by itself does offer you more TIME. Time to wait for more favourable market condition, time to make more decisons etc.

Now if you add to this, whatever edge you use as you mentioned I think it gives you an extra layer, will it work all the time ? of course not, but if I lose I take my loss just like anyone else, the downside is never any worse than the normal ways of doing business, I'm never hedging then re hedging, a loss is a loss. I just believe it does offer by itself something extra for the reasons I've given above.

Thanks for entrying in to this very interesting debate, It's been very enjoyable.

Regards.

I have likewise enjoyed the discussion.

You haven't completely convinced me of the merits of hedging, at least in terms of improved P/L, but I can see how it can offer convenience, flexibility and peace of mind, when placing one's trades.

I see the math as being like a pan-scale balance. For the unhedged trader, there are no weights on either side of the balance. The hedged trader has 1kg on each side. Now when the hedged trader commits himself to a decision, he removes one of the weights, leaving himself net long or short. The unhedged trader simply adds a weight to the appropriate side, creating an identical position, at exactly the same point. Hence the net positions wiill always be the same, as will the P/L (but the hedged trader pays additional spread).

I wish you every continued success in your trading.

All the best,
David

Hi Hanover,

I'm replying to the original post. The approach looks fine except I still don't see an edge. The positive results that you see are a result of frequent high probability wins, which is characteristic of many short gamma strategies. The problem is that this type of strategy underestimates the magnitude of the effect of that 1,000 pip loss. I used to and still do sometimes employ something similar to this technique but I establish a bias on direction and volatility to serve as a backdrop first. If you are seeing a positive return with this it could be due to a survivorship bias rather than an edge.

This is a word of caution more than anything so please don't take it the wrong way. I'm glad you had that epiphany!

If the viewers of this post are interested I recently did some simulations on systems to highlight the role that chance plays and wrote a Squidoo Lens entitled, "How to develop a sense of certainty that you will succeed in trading when the market is an uncertain phenomenon"

All the best in your trading,

Kris Matthews

Hi Kris,

I just noticed your post now. It's possible that you misunderstood my first post. I agree that there's no defined edge. All I was attempting to say was that there are alternative approaches that eschew 'conventional wisdom' that others have apparently found to be profitable.

Here's the epiphany: The bottom line is that profit is about being net long while price is rising, and net short when it's falling, frequently and/or heavily enough to overcome costs. My point was that the order types used to achieve this is irrelevant; orders are merely a vehicle to adjust net position, and it's the latter that ultimately determines P/L. That seems very obvious to me now, but it felt like a revelation at the time!

I will check out the link in your post. Thanks for sharing.

Best wishes to you also,
David

The above link seems to be out of action.

Lot's o' tasty tasty nuggets in this thread. Thanks, David.

I trust all is well down in Kiwiland. Hope the recent earthquake you had on SI didn't scramble your broadband connection. One of my sisters lives in Hawkes Bay NI and she says all is well despite the fact NZ is now about 15cm closer to Oz.

Presumably that means that AUDNZD will move just a little closer to 1.0000.

All's well up our way, thanks Opie, we didn't feel the 'quake here in Hamilton. Hawkes Bay enjoys a very pleasant climate, but I haven't been across that way since the early 1990s.

Take care, my friend.

David

I don't think this has been discussed here yet (so forgive me if it has), but IMHO assertion 1 is wrong. The selection of order type can make a huge difference to the behavior of a system, especially in the ECN environment. The difference between having a positive expectancy and not is often down to the choice of order type, especially when one is trading volatility based systems. Consider a simple breakout system, one has the choice of a market order where you could be filled at any price vs. a limit order where you know you will only be filled at a given price or better, this adds up over 1000's of trades. I understand the point being made, in that the end result is always an adjustment of exposure, but the price at which this adjustment is made is not irrelevant the way I see it.

Point taken.

All I was trying to say is that it's the net position that ultimately determines P/L, and that orders are little more than a vehicle for adjusting one's net position. For example, if I'm net short 5 lots, and I want to reduce my exposure to 4 lots, then I can either close 1 lot, or 'nedge' by opening a long position of 1 lot (assuming that my broker is not NFA-regulated).

But you're correct in saying that different order types can affect fills and therefore increase or reduce 'slippage'.

essentially i agree with what you are saying.

the question is how often you "re-balance" your position so your actual net position equals your desired net position.

the more frequent you re-balance the more trading costs in form of commissions, spread and slippage you incur.

on the other side the less often you re-balance your net position the bigger will be the volatility in your P/L which is undesirable (for example if you measure your trading performance using the sharpe ratio, or any other measure that takes volatility of the equity curve into account, your performance will be lower).
less granularity in position changes also means that the market noise will have bigger impact on the result - that is also undesirable, especially if you use some kind of optimization or data mining to adjust the parameters of your strategy.
there is also a possibility, that you will have a lost opportunity cost if you re-balance less often - for example if your rule is that you re-balance only if the discrepancy between actual and desired position is bigger than one lot, than maybe you missed the opportunity to make some profitable low volume (lower than one lot) trades.

The point is, you should research how your strategies generate signals and if you want to act on every change in the desired position.
You should find a balance between the trading costs incurred by re-balancing and the costs caused by not re-balancing should you choose to not re-balance on every desired position change.

phuuu... i hope i made somewhat understandable what i wanted to say

Nice graphic on the weighing scales.

My opinion is that the all markets are unpredictable most of the time. The trick for a trader is to find small pockets of predictablilty where the odds of a trade are in your favour.

There was an example on USDJPY today - where a 95.40 set up as a breakout level to go long. To me it looked like you should get a pop if JPY could get there and indeed it ran to 95.65 before pulling back. What happens after that initial pop is impossible to know. The edge that I was trading was that orders would likely have built up over 95.40 which could result in a pop.

Your entry criteria must give you an edge. How you manage the trade once you have open profits is a different question. What I think is important is that you took advantage of a small island of predictability to get into the market at a low risk point.

one more things - all those systems with 1:20 risk/reward are just blow-ups waiting to happen. If you are trading with a 50 pip TP and 1000 pip STOP you are making a bet that the market is rangebound within an admittedly wide range. If the market trends hard and you are on the wrong side then you are going to blow up or you have to take the trade with so little leverage that a 1000 pip stop doesn't mean a big $ loss.

Let's be clear about something (and I mean no offense to anybody and am not referring to anybody in particular):

Your edge does not come from your entry method and exit method nor does it come from money management. Sure you can have strategies with tiny consistent gains or oen with small consistent losses and big gains, but each have their tradeoffs and one doesn't have an edge over another. This is the same thing that gets options traders in trouble- they think that a certain options strategy such as a condor has an edge over simply buying calls- nothing could be further from the truth.

Your edge must come from a strategy that somehow tells you what the market is likely to do when something happens. This signal should give you a bias (long or short). Entries and Exits and other mechanical elements are nothing more than optimizations. No one can time the market precisely and the only way you know what the right entry/exit size is is by curve fitting and that will not help you in the long run, because you're assuming that what happened in the past will continue into the future.

Don't worry too much about stops. Make them small enough so that you don't get killed if you lose, and large enough so that you don't get stopped out of your profits due to intraday unpredictability.

If this doesn't make sense I'm publishing an article on this now and it will be out next week. If I remember I'll post it here.

Hope this helps save some trouble if you are fretting too much over entries and exits.

Kris Matthews

These are really good points that resonate with my current way of thinking. To date I've struggled to find any kind of statsitical edge with my exits. I've believed that "letting profit run" provides some kind of edge only because I've read posts by vastly experienced traders like Joel Rensink (e.g. here), Peter Crowns et al that seem to suggest this. "Exit is where the money is made" is a frequently voiced cliche. However, I've never been able to prove this - at least mathematically - to my own satisfaction.

Perhaps I should explain my position. I wrote post #1 almost 18 months ago, and I'd like to think I've progressed significantly since then. I still believe in the basic ideas that I expressed, though: that orders are merely a vehicle for adjusting net position, and it's the timing in making these adjustments, relative to price movements, that ultimately determines P/L. Even back then I wasn't suggesting that FXTitan or StevieT had discovered something approaching the Holy Grail. But the fact that both claimed to be extremely profitable by using wide (or no) stoplosses suggested to me that profit could be attained by means other than using 'conventional' entry/exit/MM approaches, and I was attempting to think outside the squares to ascertain why. Post #1 was my attempt to express my conclusions. Judging by some of the responses I didn't do it overly well.

When it comes to discussions on technique, I'm no longer dogmatic. Back in the 1990s, Jack Schwager said something like "there are a million ways to profit from the market, but the irony is that they're all extremely difficult to find". However, from what I've read on trading forums, many folk tend to believe that the only possible method that leads to profit is their own one, which (human nature being what it is) is understandable, but IMO it's also an ignorant and conceited assumption. Cliches and generalizations are very hard to prove universally. Just because a trader might believe that his edge comes from his exits, for example, and he has a lengthy track record to prove it, doesn't necessarily mean that it's universally applicable. Moreover, (whether rightly or wrongly) all analysis makes the assumption that an element of past behavior is more likely than not to continue to repeat itself in the future.

I agree with Kris that every entry/exit method is ultimately a compromise between win rate and win size (R:R). Increase one, and everything else being equal, the other is decreased in like proportion. The key is to find situations where everything else is not equal, an 'inefficiency'. Like JZW, I believe that there are high(er) probability entry setups, and that areas of support/resistance (like the 'flipover' S/R in the chart that he posted) can give one an edge, along with a low risk entry (which can lead to a high RR trade). Kris, I guess you mean something similar when you say "Your edge must come from a strategy that somehow tells you what the market is likely to do when something happens." If I understand correctly, that is the basis of a potential entry setup.

However, while I believe that a strong setup can help to kick-start a trade, I don't believe it's totally responsible for the end result. I've read comments like "look at how that pinbar produced a 400 pip move", but my view is that while it might have in part influenced order placement around the entry point, a zillion subsequent imponderables ultimately conspired to generate the move, and its eventual end. The simple fact is that nobody has an omniscient view as to who the participants are at every point, and where their 'collective' sentiment lies. To whatever extent certainty is lacking, trading becomes a fortuitous activity.

Having said all of this, I have no idea how I'll be thinking in another 18 months' time: another reason not to be dogmatic. I'm reminded of Gregory of Nyssa's quote "Concepts create idols, only wonder grasps anything". Perhaps that's too ethereal to be totally apt here, but certainly once one locks oneself into a certain belief set, the learning process is compromised. For me, learning how markets operate is a fascinating journey.

Hanover,

I couldn't have said it better. That's what I meant. However, I didn't mean to say the setup is the most important part, but rather getting the bias and the time frame right. Selecting the right entry, like you say, is more or less luck.

If you have the right bias (based on what you think the market should do based on fundamental or technical scenarios), price will be more forgiving if you don't time things just right.

However, if you have the right bias but the wrong time frame, you will either stop yourself out early if your stop is on the order of an average range for a smaller time frame or you will give up a profit that represented a nice move for your time frame by holding on too long. For example, if you trade on the 4 hour time frame and you have a profit after a few days but wait for a few more weeks for the profit to develop but it instead turns into a loss, you didn't respect your time frame.

Get these two things right and you will improve dramatically as a trader!

All the best,

Kris Matthews

This is an old, but very interesting thread.

I just wanted to make a couple of comments:

1. Regarding the scale diagram with unhedged and hedged positions: I agree 100% that they are identical for forex trading. However, they are not equivalent for stock trading, due to the no short-sale except on uptick rule for U.S. stocks. So for many years, hedge funds have used the "hedged trade" series of trades to effectively short a stock. If the SEC goes after them, they just say, "no, I wasn't shorting the stock, I was liquidating my (hedged) long", when mathematically, the two sides are same pnl-wise. So in this case, "hedged" is more flexible than "unhedged". There is a special name for this technique: it's called "short selling against the box".

2. With regard to having a high percentage of small winners and a tiny percentage of huge loser(s), it's an interesting idea, but hard to estimate the parameters, as you mentioned. You are effectively trying to estimate the size of the tails of a distribution. Guess what was one of the major causes of the subprime CDO mess? You guessed it - the financial engineers underestimated the size of the tails. It is a bet against the black swan showing up at your party. Selling far OTM options or even our old friend that keeps returning, the Martingale, all fall into this category. One is betting against the black swan of a long series of losing streaks. It may be something I may try when I am 90 years old, but not now. Do so at your own discretion.

Later, I'm going to reread this thread more carefully a second time. Thx for posting and taking the time to draw those diagrams.