Knowing The Future

[hwsteele]

I would like to start a thread on forecasting the market.

I have found that trading has gotten to the point of being boring. That's good and bad. Good because my returns are much more consistent. Bad because my mind wants more to do. As a result I tend to look for new ways to forecast the market. When I find one that shows promise I start to work with it and see if it can be made into a system, or at least become part of one. If it works out then that's great. If not then I will keep looking.

Now any good trader knows that forecasting and trading are two different beasts. One is fun and boosts your ego when your crystal ball works. The other smacks you around until you learned that ego has little place in good trading. Forecasting can help your trading by giving you an idea of what might be ahead, but once the forecast is made you have to let the market tell you if you were right or not. Then you must trade accordingly.

So if anyone has a forecasting method they would like to share for everyone else to look at and try on for size please feel free to explain it.
I would love to hear about it. From simple to complex it makes no difference because you never know when you are going to learn something new.:smart:
The one thing I ask is that we keep this thread positive and not put anyone down for what they share.

 

[tomorton]

Forecasting - for me - is not compatible with trading. This sounds an anomaly to newbies but you do know what I mean.

Trade what you see, not what you think -
Trade with the trend -
Wait for the signal -
Forecasting has no place in my trading if I am following these well-respected rules. I understand it is good to be able to anticipate market action, but I am happy to confess to placing both buy and sell orders on either side of a day's trading range, on occasions when I had both bullish and bearish signals. Forecasting which would be right and anticipating that move only would be superfluous and possibly keep me out of the actual trade.

 

[bbmac]

Now any good trader knows that forecasting and trading are two different beasts.

I think you are right in the above respect. Moreover forecasting will remain as you put it 'fun' certainly for for technical traders. Fundamental traders are in a sense forecasting where the market will go next but are doing so as in the case of technical set-ups to a large extent on the basis of historical precedent/experience. In a sense the only forecasting we ever do as traders is trading our edge. When trading your edge you are simply acting upon something that has proven to deliver a net gain over any historical sample of it's set-ups. In so doing, we have no need to forecast the market beacuse we do not and cannot know in advance which of our trading edge set-up's will individually produce a gain or a loss. Forecasting is therefore irrelevant to trading an edge. Mark Douglas put's it well in his writings '..We do not need to know where the market will go next to make money.'

The above said, if you have a very high strike rate trading edge then in a sense it's set-up is acting as a fairly high probability forecasting indicator of price action, but almost any trading edge whatever the strike rate when allied to the correct money management can produce an overall gain over a sample of it's set-ups, so this is a mute/acedemic point.

Personally, other than my technical trading edge and my understanding of the overall general fundamental factors at play (socio, geo, political, economic,) I have no way of forecasting the market. Often time recently I have found that with specific fundamental data releases they cause the market to react wholly oppositely to what I would have expected based on historical precedents. October's U.s NFP report was a good example of that, whereby on a -159k read the initial big knee jerk reaction was to buy $ (certainly in my preferred instrument of gbpusd) although it did sell off after that.
This and other recent examples show that you have to put individual fundamental data in the context of the wider fundamental picture, and prevailing 'mood/tone' of the market I guess, making foreacsting ever harder.

There are times when I can call almost every twist and turn of an intraday market but that doesn't mean to say that I get what I recognise as my trading edge at each twist and turn, and even if I do, that I act on them all (that's a whole other subject.) Could this gut feel/experience of observing/trading a certain instrument, getting to know it's character and idiosyncracies be called forecasting?? maybe, but I still only act when I see my trading edge, so for me it remains fun/interesting, but largely irrelevant to my trading.

Interesting subject, thanks for posting the thread, I too would be interested to learn of others' experiences with forecasting.

 

[Glenn]

You cannot forecast the market in the way that most people would interpret that.
The market is driven by Chaos Theory, which is deterministic at any point in time, but always subject to new Initial Conditions as time passes. (The weather is a good parallel).
In other words you might be correct at a particular point in time in thinking that the market will go up for several bars in whatever timeframe. If nothing else happened then it would. But once you have entered a position on the basis of your correct thinking, an event may later occur amd spoil the show.
What Chaos Theory can tell you is the time when the next event will occur (the deterministic part) , not necessarily what will happen when it does occur, because the event could be significant major news, good or bad, or just indecision, and anything in between.
In order to use Chaos Theory succesfully, the speculator needs to determine and monitor up coming event times as time passes and decide whether to alter position size or exit or let it run once the allotted time arrives.

Now all this may sound like gobbledygook and fantasy if you don't understand or accept Chaos Theory.
On the other hand ..............

Glenn

 

[hwsteele]

Very interesting so far!
Already it has been made clear that people have different ideas of what forecasting is.

tomorton said:
"Forecasting - for me - is not compatible with trading. This sounds an anomaly to newbies but you do know what I mean.

Trade what you see, not what you think -
Trade with the trend -
Wait for the signal -"

I understand what you are saying but to me the "signal" that you are waiting for IS a form of forecasting. It is a very short term forecast but with out the "signal" happening first you would not trade. So my question to you would be what "signal"/forecast do you use in your trading.
Would you be willing to give an example or two?

My own person method of trading revolves around forecasting pivot dates and times and using price pivot points. Now allot of people may say that's not forecasting at all cause I don't really call a direction or price until I am trading at the moment. i.e. If price hits a major pivot point and I see a reversal pattern(I guess you could call it a reversal forecast pattern) then I would be on the lookout for the trend to change. If it did then I would trade after conformation.

Thank you to everyone who has taken the time to explain them selves so far and I can't wait to see what others have to say!

 

[tomorton]

Hi hwsteele -

I trade using one of 6 signals (each can be either long or short) and am basically swing trading. As a favourite example, Connors & Raschke's 3-day unfilled gap reversal - if today's range leaves a gap below yesterday's low, place a buy order just above today's high, valid for next 3 sessions. Often the market will trade back into this gap within 3 days, in which case, the order will be filled and away we go: if it doesn't trigger the order, it is canceeled after 3 days. But the key is I am not opening a position based on a forecast, the market opens the position based on its own actual price action.

I am also about to add Tony Crabel's ID/NR4 as a seventh pattern - this involves placing orders both above and below today's prioce action, if today is an Inside Day with the Narrowest Range of the last 4 sessions. There is no forecast involved here at all, all I can say is the ID/NR4 pattern forecasts a breakout, but there is no way to know which direction, up or down.

Forecasting is a seductress my friend, keep yourself pure.
;-)

 

[hwsteele]

tomorton -

Thank you for giving the examples that you have. I like trading the gap also with both the eur/usd and gbp/usd. You get two or three very low stress trades a month from that.

The following is an example of one of the setups I like to trade:

I like to "forecast" pivot days in the market(S&P500 is my favorite) by "squaring" the range. Doing so gives me the target date. I don't know if it will be a high or low or at what price the market will trade it just gives me the first thing to look for. When the day comes I calculate the main price pivots for that day as I do every day. I then use this information together. An example would be:
On 8-22-08 a pivot high was made in the SPX at 1293. On 8-26-08 the pivot low was made at 1263.
The difference is 30 points. So I look 30 calendar days into the future and make a note of it.
30 days later on 9-25-08 the market had been going up from a pivot low so I know that if today works out it should be a pivot high. Meaning the price should trade down in the direction of the bigger trend. On that day I start to watch price to see what it does. If it makes a high AT one of the main price pivots for that day I watch the market to make sue price is moving away nicely and then I enter the market. In our example price topped out at 1220.03 on 9-25-08 and started down. 1220 was one of the main pivot levels for that day. The next main pivot level below it was at 1205. I do two things with this information.
One:
Even if the the day turns out NOT to be a pivot day price will most likely trade down from 1220 to the next lower main pivot level of 1205. Knowing this I can turn the trade into a day trade scalp if need be.
Two:
I make my stop 1/3 of the distance from the1220 to 1205 main pivots, or 5 points from the price I entered at. I would have gotten in about 1218 so the stop would go at 1223. This mean that the risk to reward is about 3:1 if all works out.
When the price trades down to the next lower main pivot number I move my stop to plus one point. In this example the stop would be moved to 1217.
In this example the day after entering the trade price moved down to 1188 so the stop was adjusted to 1217. Price then moved back up to 1215 and then closed at 1213 for the day. Then next day price moved down about 100 points. Most of the trading done this way the market only moves two or three main pivot numbers for the whole swing and not 100 points in one day.

I look at the above as much like trading the gap.
If price doesn't make the gap in time you don't take the trade.
If price doesn't turn at a main price pivot level then I don't trade.
If there is no clear direction of the trend on the pivot day forecasted I don't trade.

Any who, your time is VERY much appreciated and again thank you for sharing withall of us.
I believe I will be looking into Tony Crabel's ID/NR4 when Time allows.

 

[tomorton]

Hi hw - Equating points to calendar days as a forecasting system is a new one on me. I can see that, notionally, there could be a relationship, even if non-linear - the more points difference suggests a powerful price trend, which would continue for more days (and potentially have a more powerful reversal?). But this relationship seems based more than a little on a happy confluence of unrelated factors - not least of which is the number of single points currently making up the SPX. Off the wall for my taste but surely this deserves a prize for a most original approach.

 

[hwsteele]

As much as I would like to say that a price point to time unit conversion is my idea it is a very old one that has been around for at least 100 years. I really don't know why it does work I just know it does. I have read many VERY strange ideas as to why it works as well as it does, but still have no idea myself.
It is not the holy grail by any means but I have found that it works better for me personaly than any indicator ever has.
Hope every one of you have had a great weekend.

 

[kenobi]

put up a chart and then lets see how we can forecast it. then we can talk :cheesy:

 

[hwsteele]

Hi kenobi,
This is what I will do for you:
I trust you have access to charts for the SPX. If not I like BIGCHARTS.COM. They are free and halfway ok.
I will give dates and prices and then the forecast from that and if you care you can look it up to see if I'm telling the truth. Oh by the way I have nothing to do with BIGCHARTS I just like to use them when away from my computer. I like free! :clap:

This is all the SPX

July 28 low 1234.37
July 31 high 1284.93
difference 50.56 we will round up to next whole number and call it 51.
51 calender days is September 20

July 31 high 1284.93
August 4 low 1247.45
difference 37.48 We will round down to 37
37 calender days is September 10

August 4 low 1247.45
August 6 high 1291.67
difference 44.22 round down to 44
44 calender days is September 19

August 6 high 1291.67
August 8 low 1262.11
difference 29.56 round up to 30
30 calender days is September 7

August 8 low 1262.11
August 11 high 1313.15
difference 51.04 round down to 51
51 calender days is October 1

That's 5 forecasts in a row from pivot highs to pivot lows.
You decide how they did.
Again the idea is to get a general idea of when the market might pivot.
Not the holy grail but does work well for me most of the time!
I wish you the greatest success possible.