Take the 2-minute tour ×
Quantitative Finance Stack Exchange is a question and answer site for finance professionals and academics. It's 100% free, no registration required.

Many academics argue that end-to-day trading, where you go long or short before opening and sell your security at the end of the day, is not profitable. Various explanations are given for this concern. For instance,

  • macro-economic news during the day is likely to change the course of a share that day

  • in that sense, there is too much noise

  • due to algorithmic trading, the market is efficient making all profitable profits dissapear.

Still I wonder whether there exists evidence proving these statements wrong. What fo you know or think about this subject? Any good references or articles/reports on it?

share|improve this question

closed as not constructive by Tal Fishman Jun 21 '12 at 22:16

As it currently stands, this question is not a good fit for our Q&A format. We expect answers to be supported by facts, references, or expertise, but this question will likely solicit debate, arguments, polling, or extended discussion. If you feel that this question can be improved and possibly reopened, visit the help center for guidance.If this question can be reworded to fit the rules in the help center, please edit the question.

1  
In short, papers discussing if the market is efficient at a daily frequency. –  SRKX Apr 14 '12 at 11:51
    
Most papers are related to 1. modelling daily volatility 2. or are "out-dated" and correspond to my statements above. My question is whether you know a good counterargument or reference proving it wrong? –  JohnAndrews Apr 14 '12 at 12:19
4  
Many academics argue that end-to-day trading ... is not profitable. citation needed –  chrisaycock Apr 14 '12 at 15:12
    
Many academics will say all sorts of crazy things. I, for one, have never seen anyone claim that trading from open to close is not profitable but trading on either longer or shorter horizons may be profitable. It seems like a very arbitrary distinction. –  Tal Fishman Jun 21 '12 at 22:08
add comment

1 Answer 1

up vote 0 down vote accepted

Theoretically (EMH), No trading or active management is profitable consistently over time as all of those opportunities have already been exploited.

In practice: If you reduce the problem to a gambling problem.

  • Equity of K
  • Txn cost of C
  • Stock S
  • with Price at each point of time Pt
  • X - Quantity of Shares Bought
  • Target Threshold = TT
  • Target Price = Pt + TT
  • Stop Threshold = ST
  • Stop Price = Pt - ST

Case 1. Assume C = 0, TT = ST Enter at at Price Pt (Long at Pt) Max win = TT Max Loss = ST

Risk to Return ratio = 1 The probability that makes this gamble a fair bet is 50% (prob that each side is hit first).

Case 2. Assume C = C TT= ST

Max win = TT -2C Max loss = TT + 2C

Risk to Return here is (TT-2C)/(TT+2C) < 1 for any positive C

So your probability to win needs to be large enough such that p(win)(TT-2c) -((1-p(win))(TT+2C)>1 which simplifies to 2p(win)TT-TT-2C > 1 Recall P(win) here is bounded by 1. Certainty Case P(win) = 1 2TT-TT-2C > 1 TT-2C > 1 This must hold for profitability, this long expression implies that when you win you get The threshold less transaction costs.

Now to solve for the probability that which the game is fair 2p(win)TT-TT-2C = 1 p(win)= (1+TT+2C)/2TT

The probability of you being right MUST be greater than (1+TT+2C)/2TT. or 1/2 + C/TT +1/2TT Where C/TT is the ratio of cost to win. So the marginal probability over 50% in a 1 to 1 bet is equal to the ratio of Cost to win, + 1/(2*TT)

Some toy numbers from IB. if C = 2.50 (fx trade cost)

3.50+TT/2TT Becomes, 1/2 + 3.50/2TT So the probability of you winning is equal to 1 half + 3.50/2TT, where 2TT is the threshold.

As you see here you will need to be more accurate than 50%, assume that TT = 10$ 3.50/20

gives you a "fair game" probability of 67.5%, meaning to actually make money you need to be more than 67.5% accurate. An interesting fact is if TT was 175$ 1/2+3.50/350 = 51%, however the market probably does not move far enough to generate these types of profits per trade as the WIDER your target the closer the probability goes to 50% assuming fixed transaction costs.

So The lesson from this is if you aren't accurate, Have a WIDE target (in a 1 to 1 bet) as your chance of being wrong is lower, however your losses as a % of equity can be higher, given that people are more risk averse and do not have infinite capital the MAX DD's may not be tolerable.

Now with the specific case you mentioned of day to close. 1/2 + C/TT +1/2TT

The level of gains possible must be unrealistically high relative to transaction costs for it to be profitable.

Therefore unless you have a serious edge over the market, chances are you will not be profitable.

share|improve this answer
    
Could you please clarify what any of this has to do with the question, which is about the lack of profitability of trading over a very specific horizon (open to close). –  Tal Fishman Jun 21 '12 at 22:12
    
By solving the general horizon problem, with the relationship of cost to threshold explained, it should make the answer to the problem very clear.I will edit my post to clarify. –  KKB Jun 22 '12 at 17:29
    
You state at the end that a profitable trading strategy with a horizon of 1 day or less is "unrealistic." Yet there's plenty of real high frequency traders making money. Your reasoning does not make any sense to me. –  Tal Fishman Jun 22 '12 at 17:54
    
Im saying without a serious edge. I personally trade algorithmically with holding times in the minutes and without an edge I would be losing money as assuming equal win to loss threshold. Secondly it is the ratio of profit threshold to transaction cost, the higher this ratio this less accurate you can be. –  KKB Jun 22 '12 at 18:17
add comment

Not the answer you're looking for? Browse other questions tagged or ask your own question.