16 Apr 2018

Should water utilities insure against drought?

MH emailed this awhile ago:
David,

I's emailing in part due to your earlier ideas on the potential role of the insurance industry for performance oversight and rewarding proactive utility risk management (e.g., water pipe condition).

Have you evaluated the relative merits of weather risk transfer for water utilities? By this I mean index-based parametric insurance -- like the type farmers, power utilities & other weather-sensitive businesses use.

The issue as I see it: Water utilities' economic model is challenged because 80% of their costs are fixed while 70-80% of their revenues are variable, meaning that weather-driven changes in demand are financially painful.

[This post explains how mismatched revenue/cost structures lead to revenue shortfalls when falling demand lowers revenue faster than costs fall.]

Municipalities are in effect betting on the weather ("please snow enough in the mountains, please be dry during summer & then we'll sell enough water to get the funds to maintain our infrastructure").

The issue is that bad years are very disruptive -- surcharges to customers, projects delayed/deferred, staff distracted, etc. If there were a way to minimize this disruption, there could be real value created, i.e. less dead-weight loss. For regions with sufficiently volatile weather and volumetric price risk, it could be the case that tapping insurers' capital contingently based on weather triggers could be value-creating, more than offsetting the risk premium and margin that the underwriter would require.

Reserve funds should play a role for disruptions that are fairly common. But for 1 in 5 to 1 in 10 type weather events in some regions (CO, CA, TX) the variability is so great that a really massive reserve would be required. Building and maintaining such a fund would be hard, and there's the opportunity cost of capital sitting in reserve vs. being deployed on infrastructure projects where it would generate a return on assets.

Rating agencies currently do not really assess weather risk and reward/penalize water utilities for proactive weather risk management. They note the lagging indicator of Debt Service Coverage, and very occasionally they downgrade someone (Ft Worth, TX, for example).
So MH has indeed identified a real problem at water utilities where revenue is based on the volume of water sold but costs are mostly fixed.

In my understanding, such problems are the result of managers and politicians following the advice of economists on how to incentivize water conservation by increasing the share of volume-based charges on a consumers bill (meaning that the share of fixed charges falls).

The evolution works as follows:
  1. In the beginning, customers pay a fixed charge each month to use as much water as they like. Such a scheme results in stable revenue to the utility, which helps it cover its fixed costs (roughly 80 percent of the total) for its network, salaries, debt, etc. Variable costs go up and down with use, but those costs are too small to worry about.
  2. In places where water itself is scarce, there's a need to get people to use less water, which leads to a shift of charges from fixed to variable. This system is "calibrated" so that the utility receives the same total revenue and "the average consumer" pays the same total charges, so everything is good as long as behavior is "normal and average." (It's important to remember that water utilities are supposed to "break even" or make regulated profits, which explains why total revenues are set to match total costs.)
  3. The system breaks down however if people decide to use less or there is less water to sell, i.e., in droughts where people are trying to do the right thing or utilities just can't deliver enough water to meet demand.
  4. The most common response to the ensuing revenue shortfall is for utilities to raise the unit price of water to collect the same (target) revenue from selling less water. These actions are always and deservedly unpopular.
So MH is asking if it wouldn't be better for utilities to receive insurance payouts when there's a drought, as a means of covering their revenue shortfalls. (Such a system implies utilities pay the insurance companies excess revenues that would result in a year when there is "too much" water, and consumption is high.)

Although such an insurance scheme sounds like a good way to counter-act the original sin (mismatched revenues and costs), I would prefer to undo that sin by matching fixed costs to fixed revenues and variable costs to variable revenues and adding a "scarcity surcharge" when water supplies are tight (for any reason, including drought), as my system would be far more stable and avoid the need to negotiate payments and premiums with insurance companies.

Bottom line: Many water utilities use a perverse charge scheme that unbalances their finances. Such schemes should be replaced by my scheme to protect utility finances while encouraging conservation.