Quote:
Originally Posted by **apples10** Sorry, but don't agree with you - and now is a prime example. So they have the wrong vol in their model (they have actually gone down to 5%, wheras the realised 30-minute annualised is at 10%) - so they are bidding 85 on the 13:30 hourly, when it should be 81 - so you sell, but hey, the market stays up and settles at 100 (which it will!!)- you lose,so what if you sold an expensive wrong price!!
It would be hard to do a study on quiet days and 'noise' over figures as every day is different and things like weather,sport, are all factors that enter the mix - for instance was this mornings fall overdone as most traders were watching the cricket score on their Bloomberg screens?. In fact quiet days can be the worst, as there can be manipulation near the expiry.
But defo, the Dow is an enormous factor in the pm, and makes it hard to get the edge on binary trades as it is more volatile and doesn't follow the same technical patterns of the European markets. |
Apples…
But surely in your example you are ignoring the probability factor on which the example is based. If you say that it should be 81 (lets ignore the spread at this point for simplicity) then you are saying that in 81 cases out of 100 the market will settle at 100 given all the information which is available at the current moment in time. The company however were effectively suggesting that they felt that in 84 cases out of 100 the market would settle at 100. Only one of you can be correct. If you system has the edge and you are right then it equates to the following…..
You place 100 trades over a period of time…(lets say for £20 per point)…
So you lose 81 of the trades as your system predicts…
81 x 16 x £20 = £30,780 in total losses over the course of 100 bets. (The 16 comes from the fact that the company has its market at 84 hence you can only lose 16)
However, your winners will be as follows….
19 x 84 x £20 = £31,920.
Net profit from your edge is therefore £1,140 over the course of 100 bets.
I feel that my point is therefore valid. If every time you see an ‘over valuation’ you sell it and every time you see an ‘under valuation’ you buy it then you will, over a long period of time, make money as you will slowly be extracting ‘value’ from the markets. In your example the difference is 3 times out of 100 (the difference between 81 and 84). This means that in the example give, if your model is correct, you will win 3 time per 100 more than the company are predicting in their ‘valuation’ of their instrument.
As for the ‘quiet’ and ‘noise’ times. I am certainly looking to develop some strategies to take advantage of potential anomalies which occur in pricing around certain times. You must always believe that you are cleverer than the market maker. He has many prices to make whilst you can focus you efforts on just one. Obviously from your comments I can see that you have a very keen interest in the hourly bets – I can see the potential that you’ve pointed out in those. Personally I’ve always been drawn to the daily bets as I feel that they are harder for the companies to price and also more likely to priced by a non flexible pricing model. For example, imagine its 4.35pm (GMT) on an evening of a Fed announcement – how do the company price that potential volatility into their model for the daily FTSE’s. If we place bets at that moment in time can we detect potential value in opening certain bets, for example…are the odds of FTSE being +30 or -30 higher or lower for the next day if tonight has an FOMC announcement?
Got to nip out now. Will be back later with a few more ideas. Hope none of the companies are reading all this!
Steve. |