## Basic Probability

This is a discussion on Basic Probability within the The Foyer forums, part of the Off the Grid category; Originally Posted by Pacmanpth By the basic laws of economics, people are going to enter in to something until it ...

Apr 28, 2009, 9:40pm   #106
Joined Jul 2003
Quote:
 Originally Posted by Pacmanpth By the basic laws of economics, people are going to enter in to something until it no longer has profit potential (by basis of driving down/up costs or purchase price).
I think we’ve all grown out of believing in fairies and EMH.

‘People’ enter into ‘something’ well past its sell-by date – all the time. Not only when there is no clear profit potential, but even when no effort has been expended to assess any profit potential.

The majority have no concept of true value nor do they even consider it as a prerequisite to proper consideration of assessing the potential of any prospective trade.

Apr 28, 2009, 10:01pm   #107
Joined Jul 2003
Quote:
 Originally Posted by stevespray Firstly there perhaps needs to be clarification on a number of points which people, when writing, might assume others automatically consider being correct or matter of fact. The matter of the ‘coin tossing’ itself is one such instance. One or two have to my mind correctly pointed out that coin tossing itself might not be 100% random due to simple human mechanics – A human who is asked to repetitively toss a coin will develop a method and possibly a characteristic which may cause a potential imbalance in the results. It’s just the same as a professional golfer or a snooker player who constantly practices his shots – his / her aim is to program some kind of ‘repetition’ which, when called upon, the player can repeat ‘at will’. As a result we perhaps need to clarify that in reference to a ‘coin toss’ we mean a method of deriving a result which is one of two possible outcomes which has no bias one way or the other (or does that contradict what’s already written?) I guess what I mean is that we are trying to assume that there is no human influence over the ‘coin toss’ outcome?
Yeah, I don’t want to get hung up on the mechanics of coin tossing. Even the choice of coin tossing was perhaps inappropriate as I mentioned in a previous post, as most seem to think they know what probability is all about based on the elementary stuff they’ve done on coin-tossing and dice-rolling. Maybe I should have cut straight to the chase and initiated the thread based on the originating aspect for me which was futures options pricing anomalies on commodities. But then, I figured we might get stuck in that specific and narrow area of trading rather than the wider aspects of subjective probability which is where I wanted to be headed. This issue impacts all trading – not just the initiating area.

Quote:
 Originally Posted by stevespray If what I have presumed so far is true then I have to say that I am sceptical (to put it mildly) that ‘price has memory’ in terms of the tossing of coins.
I would have been too a few weeks back. But I hypothesises that the theory we currently hold to be true is only a theory. A convenient theoretical theory true, but one that increasingly shows itself to be inaccurate in practise in areas related to trading, at least.
Quote:
 Originally Posted by stevespray Of course I am willing to learn a new trick here!
You are in the minority unfortunately.
Quote:
 Originally Posted by stevespray A number of posts have been made which refer to some kind of ‘regression to the mean’ in terms of a longer series of coin tosses. Mathematically this isn’t correct although a regression can and does on occasion occur. If we spin our coin 50 times and find that we have 30 heads and 20 tails then this does not mean that the next 50 spins should yield 30 tails and only 20 heads – the probability is that there will be roughly 25 heads and 25 tails in the 50 spins. Statistically there is just as much likelihood of another 30 heads and only 20 tails as there is of 30 tails and 20 heads. On that basis I cannot consider that any kind of ‘price memory’ exists in terms of coin tossing.
The regression thing is much vaunted and as you say, it is achieved in reality far less often than classical (frequentist) probability theory would allow. I think the ’price memory’ thing causes problems when folk try and decide what it is that’s doing the remembering. It isn’t the coin…

Quote:
 Originally Posted by stevespray I would agree that people in general under estimate probability in given situations – trading is clearly one of those areas. As a result loss occurs.
Understatement of the millennium. People are incredibly arrogant in their expectations of what probability ‘means’ in any given situation. I’m going to pose a question later in this thread that if enough respond honestly (I can list now those that wont but will pretend they did) will show just how far probability does not serve those who feel they have sufficient grasp of it based on their current understanding of classical probability theory.

Quote:
 Originally Posted by stevespray Having read the thread I am having trouble in relating ‘coin tossing’ directly to trading. The thread has made me think about certain aspects of trading, and indeed coin tossing, more deeply. The more I think about it the less relationship they appear to have to each other. Initially it was suggested that trading was a matter of simply predicting which way something would move pricewise. My 10 years plus of experience has taught me that trading is so much more than that.
As mentioned. It was a bad choice with which to initiate this inquiry. Apologies.

Quote:
 Originally Posted by stevespray Randomness of price movement is a seriously interesting area to examine. I generally find that the smaller the timeframe the more random the price action is. This I would speculate is because short term price movements are caused by day to day ‘technically’ based buying / selling where as the larger time frames (weekly / monthly) are dominated by more ‘fundamental’ based buying / selling.
I wouldn’t disagree.
Quote:
 Originally Posted by stevespray I therefore find it quite hard to compare any kind of price movement to a series of ‘coin tosses’. Whether or not coin tosses are random or not does not seem even remotely significant to me. My feeling is that market movement is not random.
What is it then?

Quote:
 Originally Posted by stevespray That is not the same as saying that market movement is totally predictable. I do however have a number of ‘tools’ which I feel allow me personally to understand why certain things happen to prices. I study two forex markets fairly closely – EUR/USD and GBP/USD. Obviously these two markets are driven, in the longer terms, by completely fundamental news flow. However, in the short term, this isn’t always the case. One only has to examine the size of intraday swings to realise that these swings clock up tens if not hundreds of more pips in terms of size than the fundamental moves if studied on a weekly chart. This can only be caused by short term speculation and in my opinion this apparent randomness can be quantified if studied closely enough.
If they are driven by fundamental news flows, what moves them when there is no related news? If it is, in the shorter timeframes as you suggest, mere noise, that noise would surely be random – and therefore largely cancel out. But it doesn’t. Even when there is no direct news the shorter timeframes will develop significant (probabilistically) moves which are rarely resolved or regressed to the prior mean.
Something moves the price quite interpedently of fundamental news flows.

Nice post Steve – appreciate your thoughts.

Apr 28, 2009, 10:13pm   #108
Joined Jul 2003
Quote:
 Originally Posted by trendie I am really curious as to whether TheBramble feels he has taken this thread to a satisfactory conclusion.
LOL. I don’t think I expected a conclusion. I was asking a genuine question trendie and hoping to get some reasoned responses and on-going discourse.

To be fair, there have been a number of good and sensible responses. There have also, unfortunately, been the fairly predictable responses from the same predictable bunch…

I know from your comments in the past of your own frustrations in expectations of deeper/wider discussion on some fairly ‘out of left field’ topics – but mystical ju ju or not – you don’t get the quantity or quality you expect from a site with this size of membership of apparently experienced traders. Perhaps another issue with probability? LOL

Quote:
 Originally Posted by trendie It started off with some simple ideas/examples, then moved to abstractions and concepts. And now this thread appears to have stalled. Is this it?
trendie, there’s no point making money if you don’t take time out to spend some of it now and again, is there.

And if this (or any other thread) is dependent upon the thread initiator to keep it going, then it’s as well to let it die a natural. Either a topic is up for discussion or it’s not. If my short break from the site has been the cause, well, fair enough, I’ve responded to those posts since last being here and if there’s more mileage – let’s go for it.

Apr 28, 2009, 10:18pm   #109
Joined Jul 2003
Quote:
 Originally Posted by stevespray Whilst positions must be given room so you don’t close the position because of ‘noise’ it is also very important that the reason for being in a particular trade is constantly being reviewed. Think about this very carefully – the reason for remaining in a trade should (if you think hard enough or long enough) have NOTHING to do with the current amount of loss or profit which that trade is showing (assuming of course that you are not risking a silly %age of your trading capital on this one trade)
Is the probability the same at the start of the trade as halfway through the trade?

Of course, you don’t know you’re halfway through (time) the trade. You don’t know the trade is going to run to target and net your profit for you.

Scenario1. It’s motoring against you – hard. You’re within 2 pips of your stop.

Senario2. You’re within 10 pips of your target and it’s beginning to stall.

Your trade is going to come good, but you don’t know this. What is your estimation of the change in probability of your trade at this halfway stage compared with the probability at trade initiation?

A point Calinor touches on in a previous post. Very bright lad.

Apr 28, 2009, 10:22pm   #110
Joined Jul 2003
Quote:
 Originally Posted by TheHoneyMonster The great thing about trading is you have a 50 50 chance of picking the right direction. Pretty damn good odds - i'd say. This by no means you will make money on 1/2 the trades though, once costs have been accounted for.
No, you don’t.

Costs are an issue I agree. But the issue of direction is not a 50/50. Timing and duration are far greater issues. Capital management is an issue. Risk estimation and management are issues.

Not sure if you think the 50/50 issue is relevant to this thread based purely on the coin-toss example, but if you think trading comes down to this level of simplicity – you’re not going to be in the game for too long yourself.

 Apr 28, 2009, 10:26pm #111 Joined Sep 2001 "you’re not going to be in the game for too long yourself." Too true - honeymonster aka JTrader has been banned __________________ Mr. Charts Full time US Share Trader Follow me on Twitter @MrChartsJoyson Contributor to T2W's Trading FAQ Book http://www.trade2win.com/books/trading-faqs T2W interview http://www.trade2win.com/knowledge/a...ichard-joyson/
 Apr 28, 2009, 10:32pm #112 Joined Jul 2003 Classical (Gaussian/Frequentist) Probability Does Not Serve Us Whether you seriously and honestly attempt this exercise is up to you. I’ve got a fairly good idea of those who might respond, and of those, who will have done so with good intent and those who will likely have not. But regardless, let’s see where it goes. There’s no gold star or prize and no dunce’s hat for this one. The object of the exercise is not in the answers you may give. I’ll give it a few days to make sure everyone is really bored before I disclose the purpose and results of the exercise. Many of you may already be aware of its purpose and likely results - I’d appreciate you not disclosing it. This is a genuine test so please do not google the answer or research it anywhere else. You’re not expected to know the ‘right’ answer. I really want your ‘best guess’. Based on hunch. Select a lower and an upper range that you feel would, with a 98% probability of success, capture the highest recorded speed of any military aircraft, either in prototype or in operational duty, of any country in the world. So what range of speeds (lowest and highest) in mph do you consider will with a 98% probability, bracket the speed of the fastest military aircraft in the world?