It has been a perennial problem in the utilities industry: the weather’s unexpected impact on earnings. And in the most recent earnings season, the weather’s effect on industry financials was on full display, with most utilities reporting earnings that were largely driven by local weather patterns.

But even if we must accept Mother Nature’s overbearing demeanor, there are still other ways to identify value, and we’ll review some of those approaches below.

Historically, the weather’s role in driving utility earnings has been a source of concern for investors because it makes it difficult to discern between management’s contribution to creating value and factors that are beyond their control.

Take, for example, Duke Energy Corp (NYSE: DUK), which reported that second-quarter earnings jumped 80 percent, prompting management to increase its earnings outlook for the year. Duke’s regulated utilities benefited from higher rates and warm weather, and the firm managed to cut costs significantly. Yet, the prior year’s quarter had milder weather, so it’s hard to know how the next two quarters will stack up against comparable periods.

Then there’s CenterPoint Energy Inc (NYSE: CNP), whose regulated electric utilities reported higher revenues from customer growth, but with the caveat that “this increase was more than offset by milder weather, higher operating and maintenance expenses and higher depreciation expense.”

Although the utility is experiencing flat demand, the firm’s weather-normalized residential sales grew 2 percent over the trailing 12-month period that ended June 30, as it’s adding customers at a faster pace than many of its peers. But the question remains whether such growth can offset the weather’s potential drag on earnings.

Naturally, utilities’ management teams are not oblivious to this problem. Some firms use financial hedges against unexpected weather, while others have developed highly sophisticated weather-modelling software that they use to predict forward earnings.

And still other utilities develop business diversification strategies that they use to help smooth earnings between operations across regions with differing weather patterns, as well as varying population and demand profiles.

Eventually, greater adoption of connectivity to households through real-time metering and other smart-grid technologies will allow utilities to know more about system dynamics than ever before–and how to better control costs.

Add to that a constructive regulatory environment, and weather starts to diminish as a concern for utility investors.

But even though management teams avail themselves of many of these tools, investors don’t enjoy these same capabilities, and that means there will still be earnings surprises.

In a private meeting with a major utility’s chief financial officer during last year’s Edison Electric Institute Financial Conference, I asked whether the firm could share the results of the weather models they use to give earnings guidance.

His answer was that given the uncertain nature of the forecasts themselves, which can sometimes produce wide earnings forecast ranges, the firm wouldn’t want to make a habit of sharing the results with an investor class that demands predictable earnings.

So what is an investor to do? How can we be certain that a bad quarter was really a result of unexpected weather and not poor management?

The answer is to focus on long-term performance, which reduces the effect of factors beyond management’s control. Additionally, investors must also identify those markets and service territories that have shown superior growth in what has been an uneven recovery in electricity demand as a result of the sluggish economic rebound.

By the Numbers

To learn why some utilities performed better than others, we reviewed granular data on a state-by-state basis and discovered that utilities continue to trade very much in line with forecasted state-by-state gross domestic product (GDP) growth levels, population growth, and future climate change forecasts.

This is a continuing trend that we first observed around the same time last year, when we reviewed the US Bureau of Economic Analysis (BEA) annual release.

Before proceeding, it should be noted that the relationship between the US economy and electricity demand is changing: Growth in electricity demand has been significantly slower than GDP growth for decades. That’s why other metrics were used to serve as a counterbalance, to more effectively target utility growth.

According to a Texas A&M study published last year that forecasts energy consumption through 2050, population growth will be the single greatest factor in determining long-term trends in energy demand.

The model projects an overall national population increase of almost 38 percent from 2010 to 2050. The states with the largest growth rates included Arizona, Nevada, Florida and Texas.

Additionally, the model also used IPCC Regional Climate Projections to determine the change in energy consumption as a result of climate change in 2050.

Heating and cooling degree days for each state in 2010, as well as the average temperature for each month in that state, were provided by the National Oceanic and Atmospheric Administration (NOAA).

States with a combination of high population growth along with a large increase in the number of cooling degree days were determined to be Nevada, Arizona and Florida.

Looking at GDP, according to BEA’s 2013 data, US real GDP increased by 1.8 percent. Growth in real GDP accelerated in the second and third quarter last year after increasing at an annual rate of just 1.1 percent in the first quarter.

The BEA found that after reaching a high of 4.2 percent in the third quarter, growth in real GDP decelerated to 2.8 percent in the fourth quarter.


Real GDP grew steadily in 24 states through all four quarters of 2013. In the fourth quarter, real GDP increased in all states except Mississippi and Minnesota.

Nondurable-goods manufacturing was the leading contributor to growth in 31 states in the fourth quarter, and it consistently led growth in Louisiana, North Carolina, and Texas through all four quarters last year, according to the report.

And certainly, GDP growth figures seem to inform why some utilities are doing better than others in terms of earnings.

Of course, beyond analyzing GDP, population and growth figures, there is no substitute for looking at a firm’s long-term performance. That’s why, in 2005, I helped create an industry model for the utility journal I used to edit that could filter out those factors that are beyond management’s control.

In 2013, I developed a variation of this DuPont financial model, known as the Utility Forecaster Early-Warning System. This Early-Warning System deconstructs a company’s return on equity (ROE) into its individual components, which allows for greater ease in analyzing what’s actually driving growth.

While the Early-Warning System was developed to identify short-term improvements or weaknesses in a utility’s financials to gauge a company’s ability to sustain its dividend, the model also yields insights into management’s contribution to earnings.

Subscribers get to see the full article, which details what the model shows with regard to management’s performance for two utilities that we track.