Thursday, March 22, 2018

What draw rate should we use with a POMV approach?

One of the issues that has received some, but relatively little attention compared to others during the current fiscal debate is what the so-called draw rate should be if we start using a percent of market value (POMV) approach to establish the flow of cash from the Permanent Fund.

There have been various proposals. SB 26 (the Senate's version of a fiscal plan) uses 5.25% for a couple of years at the start, then reduces it to 5%. The Governor's most recent proposal uses 5% from the start. The House's proposed Constitutional Amendment (HJR 23) was originally 5%, but in the course of redrafting was lowered to 4.75%.

The rate used makes a difference.  The Permanent Fund has a current market value of roughly $60 billion. A draw rate of 5.25%, results in an annual draw of roughly $3.15 billion; a draw rate of 4.75% in an annual draw of roughly $2.85 billion, $300 million less. 

If split 50/50 between the PFD and government (as it should be), the use of the lower draw rate translates into $150 million less to the PFD, or, using last year's number of recipients, roughly $250 less per PFD. 

A lower contribution to the government also means increased pressure for the use of other options, such as taxes or a PFD cut.

The difference between a higher and lower draw rate increases with the size of the Permanent Fund. At a market value of $80 billion, for example, the difference between a 5.25% and a 4.75% draw rate would be $400 million, compared to the $300 million difference at a market value of $60 billion.

Generally speaking the draw rate is intended to equal the so-called "real" rate of return anticipated to be generated going forward from the investments made by the Permanent Fund. The "real" rate is equal to the overall rate of return, including inflation (the so-called "nominal" rate of return) anticipated to be generated by the Fund, less the rate of inflation.

Using the "real" rate of return is simply another (and in our view, better than the current) way of inflation proofing the Fund.

To this point in the discussion, all of the numbers have been based on the anticipated real rate of return -- that is the real rate of return anticipated to be earned in the future. That requires projections not only of the future overall rate of return that will be earned on the Permanent Fund, but also future inflation rates as well.

Generally speaking, the reason that projections are used under the POMV approach is because the goal of using a real rate of return is to keep the relevant fund protected against inflation into the future.  Thus, projections of future rates (both of return and inflation) are perceived to be more relevant than past experience.

But there also is a problem with using projected returns and inflation rates.

In a nutshell, that problem is that the future may not turn out as anticipated. Overall returns may turn out to be more or less than projected, and inflation rates may turn out to be more or less than projected.

For example, rather than the 7.5% overall return and 2.5% inflation rate anticipated by those who recommend using a 5% draw rate, the future may result in an 8% overall return and a 2% inflation rate, or a 7% overall return and a 3% inflation rate.

In the first case (8% overall, 2% inflation), a 5% draw rate will understate the 6% draw rate that actual experience demonstrates could have been used, leaving money "on the table" (i.e., in the Fund) that could have gone to support the current generation through higher PFD's and government support without penalizing those that come after.

In the latter case (7% overall, 3% inflation), a 5% draw rate will overstate the 4% draw rate that actual experience demonstrates should have been used, distributing money to the current generation that, instead, should have been retained in the Fund to protect future generations.

Normally, endowments and other institutions that use a POMV approach do not worry that much about inaccurate projections. As conditions change proving the original projection wrong, the boards of the relevant institutions simply change the draw rate to better reflect revised projections.

But the potential (even likelihood) that the future does not match projections creates a special challenge for establishing a draw rate for the Permanent Fund.  Especially if the draw rate is "constitutionalized," as proposed by HJR 23, for example, it may be hard, if not impossible, to make adjustments as conditions change, locking in permanent over or under draws and with it, potentially significant intergenerational consequences.

As a result, the more we have thought about it the more we have come to the conclusion that the better way to set a draw rate for the Alaska Permanent Fund is to use past experience, rather than a projected look. 

Using actual, past experience protects against over- or under-stating the amount which each generation may draw without impairing others. If the investment policies of the Permanent Fund Corporation result in higher than the anticipated real rate of return, as they have at many times in the past, both current and future generations will share in the resulting benefit, rather than shifting all of the benefit to future generations (and potentially leaving the current generation with higher tax rates than necessary).

If, on the other hand, those investment policies result in lower than the anticipated real rate of return, both current and future generations will share in the resulting burden, rather than shifting all of the burden to future generations.

One legitimate question about such an approach is what past period should be used in calculating the realized real rate of return. Actual real rates may vary substantially from year-to-year depending on a variety of circumstances. As with statutory earnings used in the current calculation, it is better to use an average calculated over time to smooth those out.

Changes in averages determined over time also tend to be gradual, providing the Permanent Fund Corporation with more predictability about what the draw is likely to be -- a key argument behind the use of a permanently fixed rate -- and government and residents receiving the PFD also with more predictability about what their receipts are likely to be.

In the following chart we have calculated, by year, the actual real rate of return realized by the Permanent Fund in every year since its inception, along with a rolling 5-, 10-, 15-, 20- and 25-year average.
Two things leap out (to us, at least) from charting the results in that way. The first is that the longer the time period over which the average is calculated, the smaller the year-to-year changes and the smoother the adjustments. 

The second is that every measure -- 5-year, 10-year, 15-year, 20-year, 25-year and full life (some 40 years) -- averages out to an actual, real rate of return of somewhere in the 6+% range. To us, the second point is important and raises significant questions about the substantially lower percentages (4.75 - 5.25%) that have been proposed in the current round of discussions for use in connection with the Permanent Fund.  

Recall that the lower the percentage, the lower the PFD and government draw (and the greater the potential need for finding other sources of "new revenues"). Given the intergenerational issues resulting from using a draw rate significantly lower than the actual real rate of return, to us these results add even more weight to our proposal to use a draw rate based on the actual real rate of return realized over some period of time, rather than one based on anticipated future real rates of return.

Because it generates the least amount of change year-to-year, we believe the 25-year average is the best approach, but likely could be convinced as well of the suitability of the 15- or 20-year averages. Looking at the results, we are concerned that the use of 5- and 10-year averages could generate an unnecessary amount of year-to-year change.

But at the end of the day we believe that the use of some sort of averaged actual rate of return is more appropriate than the use of anticipated rates.  The potential for significant intergenerational problems resulting from the use of a constitutionally or even statutory fixed rate is too great.