Managing Prepayments in ALM: Rate Incentive Vs Macroeconomic View

In years since the Financial Crisis, with record low and fairly stable interest rates, prepayments have not been as important as they are today.

I recently found an article from the 1990's, in a portfolio management journal, which modelled prepayments based on rate incentive for retail mortgage backed securities (RMBS). This was so that they could work out the value of the RMBS product by using the expected prepayment rate based on forward interest rates.

I back tested the rate incentive model and found it was very closely correlated with customer behaviour today, despite being a methodology from decades ago. Human behaviour seems to be consistent throughout time.

What this means is that customers are highly incentivised to prepay their mortgages early when they are on a higher rate product in a lower rate market (and vice versa). Therefore when rates increase, as they have recently, we would expect a much lower level of prepayments for customers on the back book, i.e. with much lower rates than the current market rate.

The theory being that if a customer has spare cash, they would be much more likely to invest it in a higher yielding savings product, than make overpayments on a comparatively low rate mortgage. Being on a low rate mortgage also decreases the chance of a customer repaying the mortgage in full, to move to another product/rate, because they are on a significantly better rate than the current market.

Looking at the past few years, customers were incentivised by mortgage deals of below 1% in the market in 2021, and we observed significant levels of prepayments when mortgage rates were this low.

However, from a macroeconomic lens, we might question whether this is due to rate incentive, or government intervention in the form of quantitative easing, stamp duty holiday, etc.

Personally, I do not believe it matters either way. Rate incentive and macroeconomics are very closely linked.

When the government is trying to stimulate growth in the economy, they do so in the form of quantitative easing (pumping money into the economy) and low interest rates, to facilitate spending.

Therefore, interest rates are lower and financial institutions have more cash to lend in the same instance as the housing market will be stimulated in the economy.

As there is more cash and spending in the economy, people feel more wealthy as quantitative easing leads to inflation, which increases the equity in their home and leads to a greater desire to move house. When customers move house, they mostly prepay their current mortgage deal early and take out a new mortgage for their new property.

When the government wants to limit spending and decrease inflation, the increase in interest rates will have an effect on customers which is twofold. Less spending in the economy will negatively impact the housing market and reduce customer desire to move house, but the higher interest rates will also have a rate disincentive effect, where customers are less inclined to repay or make overpayments to their low-rate mortgage in a high-rate environment.

So whether the rate incentive approach or macroeconomic approach is used to model prepayments in a financial institution, the two go hand in hand.

Although, whichever method you use to model prepayments, it will always be wrong. It is all about trying to get as close as possible. However, if your current approach is to project forward the prepayments of the prior few years into the future, it is definitely worth considering an alternative approach in today's market!