Quantitative Finance Stack Exchange is a question and answer site for finance professionals and academics. It's 100% free, no registration required.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I am attempting to recreate the S&P Dynamic Asset Exchange using the methodology outlined in this paper.

I am struggling to 'normalize' the prices of the assets properly. On page 6 of the aforementioned paper,

Price A(t) = Price of asset A normalized to equal 100 on the last trading day of the preceding year

Price B(t) = Price of asset B normalized to equal 100 on the last trading day of the preceding year

-- What methodology is implied for 'normalize' -- the standard Random Variable normalization? (Random Variable - Mean)/(Standard Deviation)? Or is there an alternative method?

share|improve this question
up vote 1 down vote accepted

I think in this case no fancy normalization techniques are implied. At least from what I understand from the cited part, they just scale the variables so that they are equal to 100 in the base period (end of preceding year) - something like computing a deflator, commonplace in macro analysis.

share|improve this answer
Hi Karol, Thank you for the answer. So, just to verify what you are saying: if the final return on the last trading day of the year was equal to 500, I would subtract 400 from all of the data points going forward? So, 12/31/2010 return is 500 -- I would subtract 400 from the return for the next trading day, say, 1/1/2011? If this is what you are describing, I have actually tried this and I have run into a number of problems. Primarily, I have run into negative returns (after they have been 'normalized'), which in turn causes problems at the ln(Price B / Price A) -- ln is undefined <0. – user2188190 Jun 24 '13 at 12:30
Are you talking about returns or price levels? Your question and comment differ in that regard. And no, I had another procedure in mind - just try to adjust the variable through multiplying by a scaling factor. E.g. $$P_{t_0}=25 \wedge P_{t_{+1}} = 30 \Rightarrow P_{t_0}^*=100 \wedge P_{t_{+1}}^* = {100\over25} \times30=120$$ At least that's what I think they meant by normalizing. – Karol Piczak Jun 24 '13 at 14:12
Thank you, Karol! Your answer seems reasonable to me -- and, importantly, results in valid data points (i.e. no negative data). Approaching this from a statistical background, in a problem that uses correlation, covariance, std, I assumed normalization in the sense of fitting the data to a normal distribution. Regarding Returns vs. Price Levels: it is my understanding that the 'Price', in this particular formula, is a function of the daily returns of various indices -- in my case, I am using the S&P 10 Year U.S. Treasury Note Futures Total Return Index to derive my Price B. – user2188190 Jun 24 '13 at 14:43

On page 6 of the document there is the formula "Ct_DeltaAB = Equation #1 evaluated for changes in Price A and Price B". How can I know what the changes are in Price A and Price B? I know the correlation between A and B but that does not mean that I can derive the changes of A and B. How do I solve this problem?


share|improve this answer
This needs to be a separate question, not an answer. – chrisaycock Jul 17 '13 at 16:02

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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