SERVING THE QUANTITATIVE FINANCE COMMUNITY

orangeman44
Topic Author
Posts: 162
Joined: February 7th, 2002, 10:13 pm

### Percentage Change

Percentage change is (P1-p0)/p0. It works fine for positive numbers, but how about negative numbers?For example,'If P1=-10 and P0=-20, what is the percentage changeWhat if P1=-20, and P0=-10?Thanks.

DrEvil
Posts: 36
Joined: April 22nd, 2004, 10:17 pm

### Percentage Change

-50% and +100%; sounds good enough for me. In one case your negative position is cut in half, and in the other, it doubles.

GregWallace
Posts: 157
Joined: August 6th, 2004, 3:31 am
Location: United Kingdom
Contact:

### Percentage Change

I have found the best formula for percentage change to be:Percentage change = (p1 - p0) / ABS (p0),where ABS (p0) is the absolute value of p0.Problems occur as p0 approaches 0 as percentage change becomes a meaningless comparative measure as it approaches +/- infinity.Also percentage change is meaningless when p0 = 0.

FedericoDyDx
Posts: 8
Joined: December 6th, 2004, 4:30 am

### Percentage Change

Another way (if you want a continuous change rather than discrete) you can take logartithms and then a time derivative (often used for, say, growth rates in an economy).Ex:Y(t)=gdp level at time 't'.log Y(t)then d log Y(t) /dt= (dY(t)/dt )/YdY(t)/dt is the time change, dividing it by Y (the whole) you obtain the percentage change.

Aaron
Posts: 6433
Joined: July 23rd, 2001, 3:46 pm

### Percentage Change

Let's try this. The example is P(0) = -20 and P(1) = -10. ln[P(0)] = 2.30 + i*pi, ln[P(1)] = 3.00 + i*pi. We don't have the values in between, but the average slope is 0.70. If you reverse the numbers, P(0) = -10 and P(1) = -20, you get the opposite slope, -0.70.

Wilmott.com has been "Serving the Quantitative Finance Community" since 2001. Continued...

 JOBS BOARD

Looking for a quant job, risk, algo trading,...? Browse jobs here...

GZIP: On