Interest Rate Forecasting Using Regression Analysis

Only available on StudyMode
  • Download(s) : 243
  • Published : February 15, 2007
Open Document
Text Preview
Interest Rate Forecasting using Regression Analysis


•Forecast of interest rates can be done in many different ways, qualitative (surveys, opinion polls) as well as quantitative (reduced form and structural approaches)* • Example of methods in quantitative approaches

- Regression method
- Univariate method (e.g. ARIMA)
- Vector autogressive models (VAR)
- Single equation approaches
- Structural systems of simultaneous equations

This paper will focus on the structural approach relying mainly on the Regression Model technique

Advantages of the structural approach:
• Rests on economic theory (unlike reduced form methods such as VAR) • Can trace the effects of changes in macroeconomic variables to interest rates (more likely long rates)

Disadvantages of the structural approach
• Data not always readily available at the required frequency

To forecast interest rates using macroeconomic variables imply the use of a structural approach – of which 2 processes are involved: 1) model building, and 2) forecasting

• Model building: model the relationship between interest rate and relevant macro variables as prescribed by economic theories and quantify (estimate) the relationships using an econometric technique • Forecasting – use the estimated model and assumptions on explanatory variables to project future values of the interest rate

Literature Review

A Structural approach to interest rate forecasting

Model building

•Economic theories: which economic variables could cause interest rate to deviate from equilibrium

•Econometric estimation: at which magnitudes would interest rates be affected by changes in economic variables (quantification of economic relationships)

Economic Theory

What is an interest rate?
•Cost of capital (price of borrowing money)
•Could be different among various players and uses: Thus different measures of interest rate (short-term, long-term, risk-free etc.)

Example of various types of interest rates

•Interbank overnight call rates
•1-day, 7-day, 14-day, 1-month repurchase rates
• 3-month, 6 month, 12 months deposit rates

Long-term (more than 1 year)
•Government/corporate bond rates
•Bank lending rates

Trend determinants of real interest rates
•Saving and Investment
•Demographic factors
•Returns to capital
•International financial integration
•Country specific risk premium

Saving and Investment
•Higher investment demand (public and/or private) relative to saving à higher interest rates •Slower productivity growth and/or slower population growth à lower returns to capital à lower investment demand •Higher consumption (public and/or private) à less saving à higher interest rates

Demographic factors
•Aging populations à "baby boomers" in the saving phase of life cycles à push up savings à but once retired à will have negative impacts on government budget and national savings

Returns to capital
•Debt competes with equity à an expected rise in return on equity will push up real interest rates •Rising productivity, economic reform, trade liberalization, lower inflation should push up business profits (at least theoretically) and thus interest rates

International financial integration
•Will lead to similar risk-adjusted returns on similar assets across countries (yet way to go)

Country specific risk premium
•To compensate for such factors as default risk, market volatility, inflation variability, etc.

Government debt

•Higher government debt relative to GDP à similar effects on long term real rates as higher current account deficit •Higher government debt relative to GDP à crowding out private investment à reduction in saving à push up long rates •Higher government debt relative to GDP à expectation of future tax increases à higher risk premium

International risk premium
Higher risk premium à higher uncertainty à higher real interest rates Can be...
tracking img