Investment Thesis

CenterPoint Energy, Inc. (NYSE:CNP) suffered significant operational, financial, and reputational damage from Hurricane Beryl in July 2024. Over 80% of CNP Texas customers lost power, and power was not fully restored for 12 days.

The stock price is down 11%, and the longer term financial and regulatory impact is still uncertain.

However, the basic investment thesis for CNP should remain intact; the demand for electricity in CNP’s Houston market is large and steadily growing, and CNP will be allowed to earn a reasonable return on the investment required to serve that market.

CNP delivers a 2.9% dividend yield today, and projects earnings per share and dividends to grow at 6-8% per year. With a share price down 11% (23% compared to the utility index), utility investors may find CNP an attractive option, particularly if the volatility induced by negative sentiment offers a slightly better entry price.

CenterPoint Basics

CNP is a multi-state regulated utility, providing gas distribution in six states, electricity distribution in Texas and Indiana, and electricity generation in Indiana. The sale of the Louisiana and Mississippi gas distribution businesses, expected to close in Q1 2025, will leave the post-sale CNP split about 2:1 electricity to gas.

Their stated goal has been to be a “premium regulated utility”. Their defining goal post-Beryl is to “build the most resilient coastal grid in the country.”

The core of the business is Houston Electric, a Transmission and Distribution Utility (TDU) which delivers electricity to 2.8 million customers in 5,000 square miles of the Houston metro area. Those customers account for 25% of the electricity consumed in ERCOT (~Texas). In 2023, Houston Electric provided 50% of CNP’s net income.

As a TDU, they own and maintain poles, wires, switches, and transformers. With the very minor exception of a few small emergency generators, CNP generates no electricity in Texas.

On 31 December 2023, their facilities included 3,910 miles of overhead and 26 miles of underground high voltage (69-345Kv) transmission lines, and 29,000 miles of overhead and 27,000 miles of underground distribution lines; about 60% of CNP customers receive power via underground lines. They own 1.17 million distribution poles. They also own 240 sub-stations, which convert high transmission voltage to lower distribution voltage.

I’ve written previous articles that described the overall CNP business, and more recently one in July 2023, CenterPoint Energy: Delivering Electricity in Texas, that provides the background to understand CNP’s role in the largely deregulated electricity market in Texas.

For the purpose of this article, I am going to focus on CNP’s Houston Electric business segment.

My Perspective

I should mention that I’ve been a CNP customer for many years, and live in the central part of the Houston Electric service area. Customers in my neighborhood are serviced by underground distribution lines, but there are above ground distribution and high voltage transmission lines nearby.

The eye of Hurricane Beryl passed near us; we lost power for 56 hours. I have family who live closer to Houston; they lost power for six days.

Six weeks post hurricane, I was awakened at 4 am by the sound of voices and equipment on the street behind my house. I walked over and found a CNP crew replacing a failed underground transformer. I talked to the crew briefly, and they mentioned they were still dealing with the after effects of the storm.

It’s been many years since I retired my last pair of muddy steel toed boots (oilfield mud, not utility), but I remember what it’s like to be out in the middle of the night, in all weather, doing work that won’t wait. The field crews have a demanding job.

How Houston Electric Gets Paid

For an investor to assess the financial and public relations issues CNP faces, it’s important to understand that CNP gets paid to deliver electricity.

The only trucks that customers see on TV repairing storm damage carry the CenterPoint logo, but CNP only gets ~ 20% of the customer’s electric bill.

In general in Texas, and specifically in the area served by CNP, the generation of electrical power and its physical delivery to customers are provided by different companies.

While CNP is a TDU monopoly, customers buy their power from a retail energy provider of their choice. Each energy provider markets directly to customers, and typically offers several plans.

The Power to Choose website lists the available energy providers and plans in Texas by zip code. There are currently 11 providers and 60+ plans in the Houston area.

Each plan details the energy charge; typically a monthly base fee, plus a usage fee, with the usage fee reflecting a fixed or variable rate of X cents per kWh used.

As an example, the default TXU Energy (Vistra Corp., NYSE:VST) plan, Flex Forward, charges $9.95 per month plus 15.2 cents per kWh used.

In addition to the cost of energy, which goes to the energy provider, there is a delivery charge, which goes to CNP. This charge in the CenterPoint area is a $4.39 per month base fee plus 3.8264 cents per kWh used.

The Power to Choose website details the costs at several usage levels. At 1000 kWhs per month, about the annual average for residential usage, an example result is shown below. The energy provider gets 79% for energy, and CNP gets 21% for distribution.

The customer gets a single bill, from the energy provider, who then passes along the TDU charges to CNP.

The numbers will vary somewhat by provider plan, but the key takeaway here is that CNP is getting only a fraction – about 20% – of the customer’s electric bill. The CNP average monthly revenue per residential customer has remarkably been ~ flat at $49 for the last 10 years.

Reliability of Power Distribution

Two key metrics are normally used to assess the reliability of power distribution:

• SAIFI = System Average Interruption Frequency Index; the number of non-momentary electric interruptions, per year, that the average customer experienced
• SAIDI = System Average Interruption Duration Index; the minutes of non-momentary electric interruptions, per year, that the average customer experienced

These metrics appear repeatedly in Public Utility Commission of Texas (PUC) documents.

Texas Utility Reliability Data

In September 2023, the PUC staff published an analysis of Electric Distribution System Spending and Reliability for Texas utilities. The analysis includes all investor owned electric distribution utilities in Texas, for the period 2013 to 2022, for Texas retail customers. One might view this as the PUC’s best effort to compare the performance of these utilities.

The average SAIFI score (outage events) was 1.81 per year (excluding one small utility which no longer exists). CNP was 2.12, ranking 6th out of eight.

The average SAIDI score (outage minutes) was 362. CNP was 222, ranking 2nd out of eight.

CNP also publishes data on reliability. Residential customers average about two outages per year, with a total outage time of 126 minutes per year (which appears to be the number without major events; see below).

National Utility Reliability Data

The US Energy Information Agency publishes an annual report on US and state power distribution reliability.

SAIFI and SAIDI data is published with and without major events (such as hurricanes), to better capture total and “normal” reliability. The “with major event” data varies considerably from year-to-year.

The table below shows the SAIDI data, with major events, for the Gulf Cost States, and for the US, from 2013-2022. Years with > 1000 outage minutes (17 hours) are highlighted in red. Note that the various between a “bad” year and a typical year for a single state can be 10x. For the US, it’s more like 2x. This begins to quantify the magnitude of the impact of an event like Beryl.

For 2022, the US SAIFI values were 1.4 outages with major events and 1.1 outages without major events. The values for Texas were 1.7 and 1.4, for Florida 1.5 and 0.8

For 2022, the US SAIDI values were 333 minutes and 131 minutes. The values for Texas were 205 and 154, for Florida 1011 and 70.

Assessment

At least for 2013-2022, Florida and Texas appear to deliver the best reliability among the states along the Gulf Coast. CNP’s performance among Texas utilities appears good, but is not outstanding. It’s likely that a Florida utility may offer the benchmark to beat for CNP’s goal to become the most reliable coastal utility.

Derecho Wind Storm Impact

Less than two months before Hurricane Beryl, the Houston metro area suffered another major weather event, a derecho, which is a widespread, long-lived windstorm. On 16 May, winds reached 100 mph in the central business district, and unusually, several large transmission towers were toppled. About 1 million CNP customers lost power.

The before and after satellite data from NASA’s Earth Observatory illustrates the extent of the 16 May 2024 Derecho windstorm on the grid in the Houston metro area. The picture on the right, taken in the early hours of the 18, is after two days of repair work.

The relevant point for our analysis here is that although the outages were less widespread, and for most customers shorter than they would see with Beryl, it significantly depleted the customer’s reservoir of goodwill toward CNP.

Hurricane Beryl Impact

The National Weather Service reported that Hurricane Beryl made landfall as a Category 1 hurricane about 80 miles southwest of Houston, around 4 am on Monday, 8 July. Maximum sustained winds were 80 mph at landfall.

The National Weather Service radar image below shows Beryl just prior to landfall. The “dirty side” of the hurricane, extends far to the east of the eye. This image provides a good idea of the extent of the storm.

After landfall, Hurricane Beryl moved north northeast, slowly weakening. By 10 am, intensity fell to a tropical storm, with the center then just northwest of Houston.

The map below shows the CNP service area around Houston (blue line). I’ve added the approximate track of Hurricane Beryl (red arrow). It was the most direct impact by the “dirty side” of a storm since Hurricane Alicia in 1983.

The post-landfall track to the north northwest meant that the eye would track slightly to the southwest and west of Houston, across heavily populated areas, and that the “dirty side” winds and rain would cover the entire CNP service footprint. The maximum wind speed was reported near Rosenberg, near the center of the CNP service area. It was about the most damaging track possible.

Rainfall exceeded 12 inches in some areas. Peak winds were measured well inland at 97 mph, and 80+ mph winds were measured across much of the CNP service area.

The damage was extensive. About 81% of CNP customers lost power, 2.26 out of 2.8 million.

We can compare that to a “normal” fair weather day benchmark. The CNP outage tracker map above also provides a summary of outages and repairs activity.

The snapshot above, at 10:30 am Sunday on 18 August 2024, provides a benchmark for the “normal” fair weather level of outages and repair activity.

At that time, there were 125 outrages impacting 2,200 customers. CNP had resolved 795 outages (33 per hour on average) effecting 13,000 customers (542 per hour on average) in the previous 24 hours.

By comparison, the extent and average duration of outages during Beryl was respectively about 1,000X and 50-100X worse than a normal day.

CenterPoint’s Immediate Response

CNP’s Houston Electric has a 91-page Emergency Operations Plan. The most recent update was filed with the PUC on 15 March 2024, i.e., prior to Beryl.

This includes an overview of how CNP organizes and manages emergency operations, and a fairly detailed process to escalate preparation activities for hurricanes based on the projected time to landfall and wind intensity. An annual hurricane drill is specified.

Restoring power after Beryl was a very big job. CNP brought in 13,000 mutual aid workers from other utilities to assist their 2,000. They removed or trimmed 35,000 trees, and replaced more than 3,025 power distribution poles. Crews physically walked 8,500 miles of lines to identify and repair damage. Mobile generators were deployed at 28 critical locations.

While they restored power to over 1 million customers within 48 hours, it took about 12 days to fully restore power (see the restoration timeline below).

Customer and Public Assessment – Negative

CNP customers were angry. A 25 July opinion piece in T&D World captured the tone:

In my 30 years of watching and covering this industry, I have never seen the same public derision and criticism of a utility that I saw with CenterPoint Energy over the last couple of weeks in the aftermath of Beryl.

The article went on to note local media headlines such as: “Why are CenterPoint repair trucks sitting in parking lots? Here’s what we found out” and “After Hurricane Beryl, why does Whataburger have a better power outage map than CenterPoint? There were online and in person threats to CNP employees.

The scope of the outages and length of the restoration process exceeded what many people were able to cope with. It was hot. Batteries on medical devices, phones, and computers ran down. In some areas cell service was degraded or failed.

Beyond inconvenience, economic damage to households and commerce was significant. Businesses could not open. People could not go to work. Refrigerated food spoiled. The economic cost of the CNP service area being without power is estimated at $1.4 billion per day.

CNP was unable to project and communicate useful expectations for when power would be restored. Their outage tracking website was overwhelmed, and reporting lacked adequate granularity.

CNP was characterized – fairly or not – as unprepared. People saw a lot of crews and equipment in staging areas, not actively doing repairs, and this was attributed to poor management by CNP.

What was a reputational and public relation problem quickly became a potential regulatory and legal problem.

On 14 July 2024, Texas Governor Abbot held a press conference, and demanded that “CenterPoint take immediate action to improve their hurricane preparation and response efforts following their repeated and ongoing power failures in the Greater Houston area”.

He also issued a letter to the Public Utility Commission of Texas (PUC) requiring them to “undertake a rigorous study to determine the causes of the repeated and ongoing power failures in the Greater Houston area following severe weather events”, and report their results by 01 December 2024, to support possible legislative action.

In addition to soliciting information from Utilities, the PUC solicited comments from the public. Over 14,000 public comments have been received.

Governor Abbot, met with CenterPoint management on 01 August 2024. A press release that day noted that “The Governor stressed that CenterPoint Energy’s proposal was inadequate and instructed them that more must be done and it must be completed faster than originally proposed.”

On 12 August 2024, the Attorney General of Texas announced an investigation into CenterPoint “for potential violations of Texas law relating to its conduct during Hurricane Beryl”.

This is still playing out, and probably will be for months.

CenterPoint’s Longer Term Response

CenterPoint had been working on improving robustness and resilience well before these storms. In February 2022, CenterPoint and the City of Houston announced the launch of Resilient Now, to develop a regional master energy plan to “enhance local power resilience across the greater Houston area.” This 10 year joint effort would include investments by both CenterPoint and the City of Houston.

CNP annually updates an existing System Harding Plan, last filed with the PUC on 31 July 2024. I found the contents unsurprising; compliance with evolving engineering standards, 5 year transmission line physical inspection cycle, 10 year pole physical inspection cycle and maintenance plan, 8 year infrared equipment inspection cycle, multi-year plan to install Intelligent Grid switching devices (80% of 1703 devices upgraded by 12/2023), increasingly sophisticated predictive failure and inspection optimization modeling, etc. It also includes projects to rebuild certain transmission lines to meet extreme wind loading and icing standards, and to elevate at risk substations to prevent flooding damage.

Beginning in 2022, certain critical distribution system poles will be engineered material rather than wood, and all freeway crossings are underground where feasible (else concrete), and all new and replacement distribution system structures will be built to comply with Extreme Wind and Extreme Ice with Ice Loading standards. In 2023, CNP installed 2,633 TripSaver devices to automatically reset distribution lines, and convert sustained faults to momentary faults; this also saves a truck roll.

It appears that CenterPoint was working the problem, but at a cadence dictated by budget constraints.

Resilience Priority Increased After Beryl

In a 24 July 2024 meeting with the PUC, CNP CEO Jason Wells said:

“Our mission is to build and operate the most resilient coastal grid anywhere in the country with best-in-class communications.”

At that meeting, CNP’s submission to the Texas Public Utility Commission, and in reply to Governor Abbot’s requests, CenterPoint outlined three major areas of focus: Resiliency, Communication, and Partnerships.

Those efforts are summarized below.

CNP reports their progress on their website here.

On 28 August, CNP provided the Texas Legislature with an update on their status and plans. This reads to me as an attempt to manage the political and regulatory reaction to Beryl. Key points:

• goal: most resilient coastal grid in the country
• double planned 2026-2028 resiliency spending to $5 billion; focus on protection from storms and extreme weather
• independent external review of resiliency cost effectiveness of investments
• Immediate Actions (July/August): trimmed vegetation for 2,000 miles of power lines, installed 300+ automation devices, installed 1,000 composite power poles
• Near Term Actions (September 2024 – June 2025): further system hardening, automation, predictive risk modeling to prepare for Winter 2024 and 2025 hurricane season; detailed workplan published by 30 September
• Longer Term Strategy (2026-2028) – to be detailed in Resiliency Plan to be filed with PUC by 31 January 2025. will include resiliency improvements for gas delivery

We will look in a little more detail at two very visible aspects of reliability – trees and power poles.

Trees Are a Major Problem

Trees and tree trimming, “vegetation management” is the term of art, were and remain a major problem. The primary cause of customer service outages with Beryl were trees falling or being blown onto power lines and other facilities. Downed trees also impeded access to power lines. Estimates are that 60% of the trees which damaged power lines fell from outside CNP’s right of way, i.e. trees that they are unable to trim without the permission of the homeowner.

The image below is a map of the CNP service footprint, divided into hexagons. The green image in on the left shows the circuit miles with vegetation near the lines. The blue image in the center show rainfall in the three months before Beryl, and the red image on the right show the duration of duration.

CNP believes that weather in recent years – high rainfall, periods of drought, and freezes – have weakened many trees. Rainfall was 2-4x normal levels in the week prior to Beryl. As the figure above indicates, areas with many trees, and with heavier recent rainfall, saw the biggest problems.

Although CNP increased spending on vegetation management by 30% in 2023, but it obviously wasn’t enough for Beryl.

If you really want the details, CNP (and other utilities) filed their 2013 to present vegetation management reports (241 pages for CNP) with the PUC as part of the Beryl response investigation.

As one might expect for a recurring maintenance task done by a large utility, tree trimming is very systematic and organized. There are ANSI standards. For 2023, about 95% of the work was contracted out to 6 large professional vendors. CNP estimated that 3,500 miles of distribution circuits are scheduled for trimming annually, prioritized by “last trim date, vegetation caused outages, potential impact on critical loads, and overall customer count impact”. About 4,600 miles were trimmed in 2023.

CNP budgeted $40 million for vegetation management in 2024, of which 81% was for scheduled proactive tree trimming, and 3.6% was for post storm activity. That’s about $14 per customer per year.

CNP’s largest challenge in this area is that contractors have a problem hiring workers qualified to trim trees around energized power lines. Public and property owner resistance to tree trimming and removal is the second-largest challenge.

Power Poles – From Wood to Fiberglass

Fallen power poles are perhaps the most iconic storm damage.

The estimated service life of wooden utility poles varies from perhaps 40 years on the low end to nearly a century on the high end. Service life varies by location, chemical treatment, and maintenance. This 2016 Oregon State University study notes that pole replacement studies find replacement rates of less than 1% per year, implying useful lifetimes well above 40 years.

CNP response reports have noted the installation of composite utility poles. An example composite pole would be Fiber Reinforced Polymer (FRP), which are custom engineered fiberglass, and claim lighter weight, twice the strength of wood, service life of up to 100 years and a sixth to a fifth the weight of concrete or steel. However, FRP poles are more expensive than wood.

In a July 10th interview, Darin Carroll, Senior Vice President of Operations at CenterPoint, said

We’ve actually already seen the benefit of some of those investments like composite poles. Even with this storm, we’ve got places down on the coast where we have some composite poles and some wood poles. The composite poles are just fine, the wood poles broke.

CNP hasn’t quite explicitly said this, but it looks like this may be the new standard going forward, at least in some areas. At ~ $10,000 per installed new pole (which is mostly installation cost), it’s $120 million to replace 1% of the installed poles.

Financial Impact

It’s too early to make a detailed assessment of the financial impact of Beryl because much of it depends on both CNP’s detailed proposal and the PUC’s response, both still pending.

CNP as a whole has a large capital investment program, $3.7 billion in 2024 ($2.2 billion in electric), and $44.5 billion ($29 billion in electric) over the 2020’s, leading to a targeted 10% growth rate in the base rate.

This has enabled CNP to target 8% non-GAAP EPS growth in 2024, and 6-8% through 2030, with dividend growth in line with EPS growth.

Storm damage cost estimates (for May and July) are $1.6-$1.8 billion.

Texas has existing statutory provisions to securitize storm damage costs over $100 million, and to recover that cost over 15 years, which is currently estimated to add about 2% to customer bills. Reacting to criticism of the utilization of temporary generators, CNP has proposed to forego $110 million of related profits.

Because of the uncertainty of the scope of system resiliency actions, CNP has elected to defer filing relevant rate requests until Q1 2025, to reflect the results of the PUC study to be completed in December 2024.

I think in the long term, the PUC will have little choice but to allow CNP to recover almost all of their costs.

CNP Stock is Down 11% Post Beryl

CNP stock price fell about 15% in the month following Beryl, then gradually recovered 1/3 of that decline. On 31 August, the stock is down 10.6%.

A comparison between the share prices of CNP and the Vanguard Utilities ETF (NYSEARCA:VPU) suggests that the impact to CNP stock is actually worse, about 23%, assuming that without Beryl CNP would have tracked VPU.

Looking back one year, CNP does appear to track VPU fairly well until Beryl hit in the first week of July.

More positively, this might also suggest that if CNP can overcome the negative reaction to Beryl, the available upside to again tracking with VPU might also be 23%. That would yield a stock price of about $33.50.

Wall Street (17 analysts) is less optimistic, predicting an 8% upside to $29.61.

With short interest reported by Seeking Alpha at 1.87%, the market doesn’t appear to see significant downside risk.

Risks to Investment Thesis

The biggest risk that I see is regulatory, i.e. political.

Based on my experience doing business performance audits, I’d be very surprised if the PUC investigation didn’t identify a number of things that CNP could have done better, and should change going forward. It’s a rare post-mortem analysis that finds no faults.

I would be surprised, however, if almost all of these problems aren’t the result of budget limitations (e.g. not spending another $40 million a year on tree trimming) and the frictions inherent in quickly responding to a crisis. Having an emergency response plan is not the same as having the ability to smoothly execute the plan. It’s expensive – in money and management attention – to seriously prepare and train for relatively rare events.

In the end, I think the PUC should allow, and in fact demand, that CNP spend more money on resilience, and set the rates to support that. For example:

• Increase the vegetation management tempo; 3,500 miles budgeted per year vs. 29,000 miles of overhead distribution lines means an 8-year cycle. Even optimized with predictive analytics, that may not be enough. I pay to have the big oak in my backyard pruned every other year, specifically to mitigate the risk from high winds.
• Authorization from the PUC and Legislature to trim/remove at least some trees outside the utility easement, with compensation to the owner.
• Install technical upgrades (Intelligent Grid Switching Devices, TripSaver breakers, anti-galloping ice/wind devices on transmission lines) and facility protection (elevating flood risk substations) at a pace limited by execution capacity rather than by budget.
• Accelerate distribution grid architecture improvements (sectioning) and component changes to support projected higher electrical loads.
• Ensure adequate stockpiles of components (poles, transformers, etc.) for emergency repairs.

However, if the politics turn punitive, the short-term attraction of being seen to punish CNP, at least to some degree, may prevail.

I hope, and expect, that common sense will prevail.

Wider Investment Implications

Investors may want to consider implications here that extend beyond CNP.

Many utilities face major investment requirements to build and then connect generation resources to meet new demand. Widely predicted increases in electrification of transportation and other will stress many utilities, regulators, and customers.

On the generation side, ERCOT produces a Monthly Outlook for Resource Adequacy Report, which provided a statistical analysis of possible future supply shortfalls severe enough to cause rolling blackouts. The report for August 2024, released in June 2024, suggested a 12% chance of a shortfall in August. That didn’t happen … this year.

Similar warnings about future supply shortfalls have been made by several other regional grid managers (MISO for next summer, and PJM longer term).

There will be increased interest at the customer level in mitigating the impact of grid reliability and generation shortfalls.

Companies such as Generac Holdings Inc. (NYSE:GNRC) that provide whole house backup generators will see more demand. Advertisements for Generac backup systems are running every day on local Houston TV.

Many residential solar systems shut down when the grid is down, but some such as Enphase Energy, Inc. (NYSE:ENPH) offer systems that will continue to function with the grid down.

Investor Takeaway

Hurricane Beryl caused significant operational, financial, and reputational damage to CenterPoint. While the immediate operational problems have been repaired, the longer term financial and repetitional impact is still uncertain.

As a result, the stock declined 15% before recovering a bit, now down only 11%. However, it’s down about 23% vs. the utility index.

My general assessment is that CNP is a reasonably well managed regulated utility, with a long runway of investments ahead to grow its rate base, operating in a growing market. Beryl will probably serve to increase and expedite resiliency investments, and internal CNP improvements.

I would still prefer that CNP divest their Indiana Electric asset, and focus their attention on their Houston TDU. The Indiana footprint is not large enough to move the needle, and divesting it would significantly simplify their business. That might, in fact, be one option to respond to complaints that CNP has too much on their plate. Perhaps one day.

Personally, I have a position acquired in multiple trances in 2015 and 2020 with an average cost of $20.58, which provides a 3.9% yield on cost. It’s my only utility, and a long-term hold for me, primarily reflecting a bet on Houston’s growth.

Under the Seeking Alpha classification system, I would rate CNP today at as Hold. However, with a share price down 23% compared to the VPU utility index, investors seeking utility exposure might find CNP worth a closer look, and perhaps a limit order at $25.