Содержание
- 2. Lecture Objectives Introduce the idea and rationale for forecast averaging Identify forecast averaging implementation issues Become
- 3. Introduction Usually, multiple forecasts are available to decision makers Differences in forecasts reflect: differences in subjective
- 4. Introduction Disadvantages of using a single forecasting model: may contain misspecifications of an unknown form e.g.,
- 5. Outline of the lecture What is a combination of forecasts? The theoretical problem and implementation issues
- 6. Part I. What is a combination of forecasts? General framework and notation The forecast combination problem
- 7. General framework Today (at time T) we want to forecast the value of (at T+h) We
- 8. Notation is the value of Y at time t (today is T ) h is the
- 9. Interpretation of loss function L(e) Squared error loss (mean squared forecasting error: MSFE) equal loss from
- 10. A combined forecast is a weighted average of M forecasts: The forecast combination problem can be
- 11. Clarification: combining forecasting errors Notice that since then Hence, if weights sum to one, then the
- 12. Summary: what is the problem all about? (II) We want to find optimal weights (the theoretical
- 13. General problem of finding optimal forecast combination Let: u an (M x 1) vector of 1’s,
- 14. Issues and clarifications Do weights have to sum to one? If forecasts are unbiased, this guarantees
- 15. Summary: what is the problem all about? (I) Observations of a variable Y Forecast observations of
- 16. Part II. The theoretical problem and implementation issues A simple example with only 2 forecasts The
- 17. Optimal weights in population (M = 2) Result 1: The solution to Problem 1 is weight
- 18. Interpreting the optimal weights in population Consider the ratio of weights A larger weight is assigned
- 19. Result: Forecast combination reduces error variance Compute the expected MSFE with the optimal weights: |ρ| ≤
- 20. Estimating Σ The key ingredient for finding the optimal weights is the forecast error covariance matrix,
- 21. Issues with estimating Σ Is the estimate of based on the past forecasting errors “good”? If
- 22. Optimality of equal weights The simplest possible averaging scheme uses equal weights The equal weights are
- 23. Part III. Methods to estimate the weights: M is small relative to T (M
- 24. To combine or not to combine? Assess if one forecast encompasses information in other forecasts For
- 25. OLS estimates of the optimal weights Recall the general problem of estimating wm for m forecasts
- 26. Reducing the dependency on sampling errors Assume that estimate is affected by a sampling error (e.g.,
- 27. Part IV. Methods to estimate the weights: when M is large relative to T
- 28. Premise: problems with OLS weights The problem with OLS weights: If M is large relative to
- 29. MSFE weights (or relative performance weights) Relative performance weights An alternative to the of OLS weights:
- 30. Emphasizing recent performance Compute: where is the number of periods with δ(t)>0 and δ(t) can be
- 31. Shrinking relative performance Consider instead As parameter k 0 the relative performance of a particular model
- 32. MSFE weights ignore correlations between forecasting errors Ignoring it, when it is present decreases efficiency –
- 33. Rank-based forecast combination Aiolfi and Timmerman (2006) allow the weights to be inversely related to the
- 34. Trimming In forecast combination, it is often advantageous to discard models with the worst and best
- 35. Example Stock and Watson (2003): relative forecasting performance of various forecast combination schemes versus the AR
- 36. Part V. Improving the Estimates of the Theoretical Model Performance: Knowing the parameters in the model
- 37. Question So far we assumed that we do not know models from which forecasts originate Would
- 38. Hansen (2007) approach For a process yt there may be an infinite number of potential explanatory
- 39. Hansen (2007) approach (2) Let be the vector of T-h (in-sample!) residuals of model m The
- 40. Example of Hansen’s approach (M = 2) We need to find w that minimizes the Mallow
- 41. Conclusions – Key Takeaways Combined forecasts imply diversification of risk (provided not all the models suffer
- 42. Thank You!
- 43. References Aiolfi, Capistran and Timmerman, 2010, “Forecast Combinations“, in Forecast Handbook, Oxford, Edited by Michael Clements
- 44. Appendix
- 45. Appendix 1: generalization of problem 1 Let w be the (M x 1) vector of weights,
- 46. Result 1: Let u be an (M x 1) vector of 1s’ and ΣT,h the variance-covariance
- 47. Appendix 2: generalization of result 1 Let e be the (M x 1) vector of the
- 48. Appendix 2: generalization of result 1 (M = 2) Let Σt,h be the variance-covariance matrix of
- 49. Optimal weights in population (M = 2) Result 1: The solution to Problem 1 is weight
- 50. Appendix 3 Notice that Need to show that the following inequality holds and that Rearrange the
- 51. Appendix 4: trading-off bias vs. variance The MSFE loss function of a forecast has two components:
- 52. Appendix 4 The MSFE loss function of a forecast has two components: the squared bias of
- 53. Appendix 5 Suppose that where P is an (m x T) matrix, y is a (T
- 54. Appendix 5 Consider:
- 55. Appendix 6: Adaptive weights Relative performance weights may be sensitive to adding new forecast errors (may
- 57. Скачать презентацию