Welcome to Dr Lincoln's blog

Welcome for visiting my blog. Hope you enjoy the visit and always welcome back again. Have a nice day!

2009-05-07

Taylor's rule enhanced and expanded

This post contains some further thoughts on Taylor's rule (see http://en.wikipedia.org/wiki/Taylor_rule), beyond my earlier posts on this issue (see e.g. http://mrlincolns.blogspot.com/2009/05/taylor-rule-of-monetary-policy-impotant.html).

Better indicator for an existing variable
Use of output. In a dynamic sense, the desired output is unknown, although one can project it using past information. One alternative is to use employment, but it suffers the same problem. Besides, use these sorts of variables with very different units to the target policy variable, that is, the official interest rate, is a bit undesirable. The use of logarithm converts these to a much manageable and smaller numbers, but their differential means the parameter to use is a little harder to choose, and possibly to interpret.

Another option is to use unemployment rate, but in a reverse order to that of the output’s. This can make the unit comparable to that of the inflation variable. This appears to be a better choice.

Use additional indicator/indicators
Having an indicator for the bounds, in which Taylor’s rule works or is expected to work at least in a theoretical sense, could make the rule more complete. My main concern about the rule’s effectiveness is that it is ineffective under some scenarios, including stagnation, great recessions/depressions during which the conventional monetary policy prescriptions do not work when liquidity trap occurs.

I think it would highly desirable to have an indicator for stagnation and an indicator for liquidity trap including the more severe case where normal banking and finance activities are seriously curtailed.

Indicator for liquidity trap. One option is to use the amount of financial institutions’ loans. The indicator could be constructed as a ratio of ratios. The first ratio is this amount to the expected normal level of output. From this ration we can calculate a multi-period moving average of this ratio. The indicator for liquidity trap is the ratio of the current period ratio of aggregate loans and desirable output, to the moving average of this ratio. If this indicator is below a certain level, say 0.9, then there is a problem of liquidity in the economy. This level could be estimated using past information.

Indicator for stagnation. One option for this indicator is similar to the construct of the liquidity trap one. In this case, however, we use both inflation and growth variable as the components for the indicator. For the inflation variable, we could use the ratio of current period inflation to its multi-period moving average. For the growth variable, we could use the ratio of the current period growth rate to its moving average.

Once we have these two ratios, we can construct another ratio, namely, the ratio of inflation ratio to the growth ratio. If this ratio exceeds a certain level, then it indicates the existence of stagnation. However, this construct has a serious problem in that the denominator may be zero or negative or very close to zero. When the growth ratio is close to zero, this indicator will be very large. Is that a fatal flaw with this construct? Need to further consider it.

With these two indicators, we can specify or estimate the conditions under which Taylor’s rule will become ineffective and will cease to be useful as a guide for the conventional monetary tool. In those conditions, other monetary tools (see ) have to be used.

There might be other situations that will also render Taylor’s rule ineffective. It is desirable to identify those conditions if there are any. Once those conditions are identified, we could further modify Taylor’s rule.

No comments:

Post a Comment