I discussed the mathematics of the PRA proposal to tie the deferment rate to the real risk free interest rate here, linking to two earlier posts by Kevin with the detailed mathematics. But we can express our result without much of the detailed mathematics, as follows.
Here is the classic dividend discount model equation.
q = r + π – g
where q is the net rental rate, r is the interest rate, π the risk premium and g the growth in rental (NB not the growth in asset price, which is completely wrong).
We can interpret r and g either as real or nominal rates. If real rates, then r is the real return after indexing to some inflation index and g is the real growth after indexing to the same inflation index.
The inflation index is typically a general index such as RPI or CPI, but we can easily use another. How about a rental index? Suppose the government issues a perpetual bond indexed to a net rental index, Then r is the risk free rental-indexed rate of return. The risk premium π is zero, because the government is issuing the bond. What about g? Well it is the growth of rental indexed to the rental index! So g is zero by definition. Then
q = r
which is exactly as you would expect! The risk-free deferment rate must be the same as the rate of return on a government bond indexed to rental inflation.
But the deferment rate on a portfolio of properties is not risk-free, and the risk premium π is likely to be greater than zero. Indeed, we can infer a likely premium from data on the net rental rate, see page 40 of our ERM Guide which suggests a rate (based on Radu Tunaru’s report for the Institute of Actuaries) of 4.2%. Thus, assuming a real risk-free rental coupon of close to zero, in line with the currently low rate of RPI bonds, we infer a property risk premium of around 4%.
Why then is the PRA assuming a minimum deferment rate of 1%?