Riskalyze analysis is data driven. Riskalyze uses real historical pricing data to calculate the volatility and correlation of an investment and portfolio. It's great when a security has a long history, but what about a relatively new security?
Extrapolation significantly enhances the risk/reward analysis of young securities by using the actual data to find similar securities and fill in the missing history with a close approximation of how the security would have likely performed during 2008 and 2009. By the way, Riskalyze will add a note any time Extrapolation is being used in a portfolio:
Riskalyze gives you superior analysis with young funds to help you reassure your clients that they are invested correctly. Extrapolation adds incredible depth to analysis while staying true to the values of objectivity and simplicity that our advisors love.
How does Riskalyze extrapolation work?
We believe that a robust probability range (and Risk Number) is best calculated using data from throughout a varied economic environment. Today we collect and calculate return, volatility and correlation data from January 1, 2008 - present. We want the data to include good, bad and ugly market environments. Investments with inceptions more recent than January 1st, 2008 require us to mathematically compensate for the holes in the data.
For young (post-2008 inception) securities, we extrapolate volatility based on the investments correlation to the Standard & Poor's 500.
For example, assume an investment that launched mid-2012 has historically demonstrated half as much volatility as the Standard and Poor's 500, we would calculate an extrapolation ratio of 2.0. That is, if you look at SPY from June 2012 to present, the calculated sigma of the young investment is half of what it would have likely experienced from January 2008-present. In this example, we would double the calculated volatility. If the 2012-present volatility was calculated as 8 we would adjust this to a volatility of 16 (calculated actual sigma of 8 x extrapolation adjustment 2 = post-adjustment volatility of 16).
If a fund's inception was at the market bottom (April 2009) we believe it actually has demonstrated about 70% of the true volatility (extrapolation ratio is 1.4: 1/1.4~=0.71), despite only lacking ~12 months of data from our desired full data set.
This methodology allows Riskalyze to 'back-fill' volatility data for investments that lack data from a full market cycle using an objective -statistically robust- approach.
Additional insights can be found on our blog announcement here.