Skip to content
All library documents

Reconstructing a Discontinued Commercial Paper Rate with Regression

Article Quant Q&A · Author: Zach
Read the original

Summary

The document proposes a regression-based method for extending a discontinued three-month commercial paper rate beyond its final observation. The first step is to identify a basket of similar interest rates with overlapping historical data before and after the discontinuation. Regress the old commercial paper series on that basket, then use the fitted relationship to estimate the discontinued rate in the later period.

Next, regress the fitted series on the two newer commercial paper rates, financial and nonfinancial, to estimate how they relate to the old measure. The resulting coefficients may be used as weights, with options to omit the intercept or normalize weights to sum to one. The answer gives a workflow rather than regression results, selected predictors, or a validated reconstruction. Its usefulness depends on overlap data and on whether historical relationships remain suitable for the period being estimated; no diagnostics or uncertainty measures are specified.

Key ideas

  • Choose comparable interest rates with data overlapping the discontinued series and later periods.
  • Use the historical basket to fit the old commercial paper rate and extend its estimates forward.
  • Regress the fitted estimates on newer financial and nonfinancial commercial paper rates.
  • Consider omitting the intercept or normalizing the estimated coefficients as weights.
  • The proposed procedure supplies no model diagnostics or evidence that the relationships remain stable.

Tags

Fixed incomeStatisticsBacktesting
Full text
# How to reconstruct a discontinued economic time series such as the Fed's CP rate?


# How to reconstruct a discontinued economic time series such as the Fed's CP rate?












The old 3-Month Commercial Paper Rate (CP3M) on FRED was discontinued in 1997. I would like to reconstruct this series in a reasonable fashion, so I can use it to analyze more recent events.

I was thinking of using some combination of 3-Month AA Financial Commercial Paper Rate (CPF3M) and 3-Month AA Nonfinancial Commercial Paper Rate (CPN3M), but I'm unsure how to weight them.

Any suggestions?

## Answer by Tal Fishman (score 1, accepted)

https://quant.stackexchange.com/a/1712

Take a basket of "similar" interest rates from the Fed's H.15 data page which have sufficient (at least a few years) data both before and after the 1997 discontinuation date. Run a regression of the old CP rate on those variables, derive a fitted CP rate, extrapolate the fitted CP rate past 1997, then regress the fitted CP rate on the two new CP rates. Use the coefficients from this regression, perhaps excluding the constant and perhaps normalizing the coefficients to be weights that sum to 1.

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.