Probability for SFratio
- Thread starter kaia_p
- Start date
S2000magician
New member
- Jun 18, 2026
- 0
- 0
Color me ignorant, but I have no idea what you mean by “SFRatio” nor its probability. If you would clarify that, I’d be happy to try to help.
S2000magician
New member
- Jun 18, 2026
- 0
- 0
OK … now I’m trying to figure out where the N and probability come into this.
Roy’s Safety First criterion is just the ratio; you compare a number of investments, and the one that has the highest ratio is the best. I don’t see how probability comes into this.
Where did this arise?
Roy’s Safety First criterion is just the ratio; you compare a number of investments, and the one that has the highest ratio is the best. I don’t see how probability comes into this.
Where did this arise?
SFRatio tells you simply whether you should invest in a security or a portfolio. It is the expected return - minimum required return divided by the standard deviation. The highest value is usually the best.
N(SFRatio) calculates the probability that the stock/security in question will have returns falling below the threshold of minimum required returns. So you assume a standard normal distribution and calculate the Z-value of each, then measure the CDF. Again, the highest value for SFR will always lead to the lowest probability of Pr(Rp<Rl).
-N(SFRatio) is the probability that the above scenario will happen. Lower the better.
1-N(SFRatio) is the opposite.
N(SFRatio) calculates the probability that the stock/security in question will have returns falling below the threshold of minimum required returns. So you assume a standard normal distribution and calculate the Z-value of each, then measure the CDF. Again, the highest value for SFR will always lead to the lowest probability of Pr(Rp<Rl).
-N(SFRatio) is the probability that the above scenario will happen. Lower the better.
1-N(SFRatio) is the opposite.
S2000magician
New member
- Jun 18, 2026
- 0
- 0
Aha!
I hadn’t realized that kaia_p was using N(Z) to signify the cumulative standard normal probability function.
The scales fall from my eyes.
I hadn’t realized that kaia_p was using N(Z) to signify the cumulative standard normal probability function.
The scales fall from my eyes.
- Thread starter
- #8
sorry still a bit confused. let me give you an example, please help to explain what the N is
For example, suppose an investor’s threshold return, RL, is 2 percent. He is presented with two portfolios. Portfolio 1 has an expected return of 12 percent with a standard deviation of 15 percent. Portfolio 2 has an expected return of 14 percent with a standard deviation of 16 percent. The SFRatios are 0.667 = (12 − 2)/15 and 0.75 = (14 − 2)/16 for Portfolios 1 and 2, respectively. For the superior Portfolio 2, the probability that portfolio return will be less than 2 percent is N(−0.75) = 1 − N(0.75) = 1 − 0.7734 = 0.227 or about 23 percent, assuming that portfolio returns are normally distributed.
thanks!
For example, suppose an investor’s threshold return, RL, is 2 percent. He is presented with two portfolios. Portfolio 1 has an expected return of 12 percent with a standard deviation of 15 percent. Portfolio 2 has an expected return of 14 percent with a standard deviation of 16 percent. The SFRatios are 0.667 = (12 − 2)/15 and 0.75 = (14 − 2)/16 for Portfolios 1 and 2, respectively. For the superior Portfolio 2, the probability that portfolio return will be less than 2 percent is N(−0.75) = 1 − N(0.75) = 1 − 0.7734 = 0.227 or about 23 percent, assuming that portfolio returns are normally distributed.
thanks!
S2000magician
New member
- Jun 18, 2026
- 0
- 0
N is the cumulative standard normal distribution: N(Z) = P(z ≤ Z), where z’s distribution is standard normal (μ = 0, σ = 1).
For portfolio 1, the SF ratio is 0.67; i.e., the minimum acceptable return is 0.67 standard deviations below the mean. If the distribution of returns is normal, then the probability that the return will be below the minimum acceptable return is P(z ≤ -0.67) = N(-0.67). From table of cumulative standard normal distribution values, N(-0.67) = 0.2514, so there’s a 25% chance that portfolio 1 will have a return less than 2%.
For portfolio 2, the SF ratio is 0.75; i.e., the minimum acceptable return is 0.75 standard deviations below the mean. If the distribution of returns is normal, then the probability that the return will be below the minimum acceptable return is P(z ≤ -0.75) = N(-0.75). From table of cumulative standard normal distribution values, N(-0.75) = 0.2266, so there’s a 23% chance that portfolio 2 will have a return less than 2%.
For portfolio 1, the SF ratio is 0.67; i.e., the minimum acceptable return is 0.67 standard deviations below the mean. If the distribution of returns is normal, then the probability that the return will be below the minimum acceptable return is P(z ≤ -0.67) = N(-0.67). From table of cumulative standard normal distribution values, N(-0.67) = 0.2514, so there’s a 25% chance that portfolio 1 will have a return less than 2%.
For portfolio 2, the SF ratio is 0.75; i.e., the minimum acceptable return is 0.75 standard deviations below the mean. If the distribution of returns is normal, then the probability that the return will be below the minimum acceptable return is P(z ≤ -0.75) = N(-0.75). From table of cumulative standard normal distribution values, N(-0.75) = 0.2266, so there’s a 23% chance that portfolio 2 will have a return less than 2%.
Thank you- much appreciated!