Question

Suppose you had a large number of variables which you believe might have some forecast information. How would you use these variables to forecast in an efficient way?

B) How would you extend this procedure to allow for dynamic effects?

Answer 1

(A) Six Rules for Effective Forecasting:-

1. Define a Cone of Uncertainty- you ultimately have to rely on your intuition and judgment.  It broadens your understanding by revealing overlooked possibilities and exposing unexamined assumptions regarding hoped-for outcomes. At the same time, it narrows the decision space within which you must exercise your intuition.

2. Look for the S Curve- The most important developments typically follow the S-curve shape of a power law: Change starts slowly and incrementally, putters along quietly, and then suddenly explodes, eventually tapering off and even dropping back down.

3. Embrace the Things That Don’t Fit- The entire portion of the S curve to the left of the inflection point is paved with indicators—subtle pointers that when aggregated become powerful hints of things to come.

4. Hold Strong Opinions Weakly- One of the biggest mistakes a forecaster—or a decision maker—can make is to overrely on one piece of seemingly strong information because it happens to reinforce the conclusion he or she has already reached.

5. Know When Not to Make a Forecast-  Even in periods of dramatic, rapid transformation, there are vastly more elements that do not change than new things that emerge.

6. Look Back Twice as Far as You Look Forward- The recent past is rarely a reliable indicator of the future

forecasting using variables:-

1. Automatic model selection techniques can be an effective and efficient method of forecasting in certain business situations.
2. use techniques such as principle components and other shrinkage techniques, including Bayesian model averaging and various bagging, boosting, least angle regression and related methods.
3. using “hybrid” combination factor/shrinkage methods often yields superior predictions.
4. To forecast a response series by using an ARIMA model with inputs, you need values of the input series for the forecast periods. You can supply values for the input variables for the forecast periods in the DATA= data set, or you can have PROC ARIMA forecast the input variables.
5. vector autoregressive (VAR) model- This sees the four variables as fundamentally stochastically linked in ways that you may not want to assume

(B) for dynamic effects:-

Time series data are data collected on the same observational unit at multiple time periods .

Example 1 of time series data: US rate of price inflation, as measured by the quarterly percentage change in the Consumer Price Index (CPI), at an annual rate

• To develop forecasting models - What will the rate of inflation be next year?
• To estimate dynamic causal effects

(a) If the Fed increases the Federal Funds rate now, what will be the effect on the rates of inflation and unemployment in 3 months? in 12 months?

(b) What is the effect over time on cigarette consumption of a hike in the cigarette tax?

(c) Rates of inflation and unemployment in the US can be observed only over time!