Choosing Equity and Bond Indices for 60/40 Portfolio Simulations
Summary
The document discusses selecting historical index data for testing portfolio optimization methods and benchmarking a 60/40 equity and fixed-income allocation. It emphasizes that the appropriate series depends on the portfolio the simulation is meant to represent: index constituents and exposures should resemble the assets an investor would actually hold. It also raises practical constraints around free data sources and the desire for a long historical record.
Suggested equity benchmarks include broad US and international indices, while fixed-income possibilities include US Treasury, corporate, and aggregate bond indices. A separate suggestion is to use the US 10-year constant-maturity yield series, though a yield series is not itself a bond total-return index. The listed histories vary in starting date and availability, and the document gives no download workflow or performance results. The index choice therefore remains use-case dependent: a simulation needs series that match its intended exposures, and price, yield, and total-return data should not be treated as interchangeable when evaluating portfolio performance.
Key ideas
- Choose simulation indices to reflect the stock and bond exposures the eventual portfolio is intended to hold.
- Broad US, global, and regional equity indices offer different coverage and historical availability.
- Treasury, corporate, and aggregate bond indices represent different fixed-income segments.
- A constant-maturity Treasury yield series is a possible data reference but differs from a bond total-return series.
- The document provides candidate benchmarks rather than a data-fetching method or a comparison of results.
Tags
Full text
# Which indices to use for an equity vs. fixed-income portfolio simulation? # Which indices to use for an equity vs. fixed-income portfolio simulation? I want to backtest several basic optimization methods (e.g. MVO, "most-diversified portfolio"), and I want to do this on a basket of different asset indexes. To start with, I want to simulate a 60/40 portfolio (will use this as a benchmark). What price series should I use, in particular, for bonds? I want to use a free, open source of data, which usually means getting data off of Yahoo! or FRED. I want a long history, back to the 80's if possible. Does anyone have an example of R code to fetch a good equity and bond portfolio and show the performance of a 60/40 portfolio? ## Answer by Ram Ahluwalia (score 2) https://quant.stackexchange.com/a/2695 On the bond side, perhaps the yield on the 10-year constant maturity series available at the St. Louis Federal Reserve (FRED website). On the equities side, S&P 500 or a global index such as MSCI are good but you might not have history thru the early 80s on the latter. ## Answer by Tal Fishman (score 1) https://quant.stackexchange.com/a/2703 The answer to your question largely depends upon how you will be using the results of your simulation. I strongly believe that you should match the index you choose for simulation as closely as possible to the actual mix of stocks and bonds in which you will ultimately be investing. Having said that, the most popular indices for these asset classes are: Equity - Russell 1000 - US large + mid cap - data since 1979 - Russell 3000 - US all cap - data since 1979 - MSCI USA/EAFE/World - International country/regional indices - data since 1970 - S&P 500 - US large cap - data since 1950s or earlier (backfilled) Fixed Income - Lehman/Barclays US Treasury - data since 1973 - Lehman/Barclays US Corporate - data since 1973 - Lehman/Barclays US Aggregate - data since 1976 I'm not sure where to get these data for free, but you should be able to get a monthly returns series at relatively low cost. See our data question for suggestions and leads.
Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.