Viewing the Inflation Reduction Act Through the Lens of Energy Infrastructure

The Inflation Reduction Act (IRA), signed into law by President Biden on August 16, includes provisions for $370 billion in energy-and climate-related initiatives, making it the most significant environmental-focused bill Congress has ever passed. The bill has the potential to impact a wide range of industries. The stimulus is meaningful; however, the caveats and stipulations around each of the bill's provisions should elicit caution in making any concluding statements on the eventual impact.

The most transformative areas of the bill appear to be:

1. Battery storage, which will be eligible for a 30% investment tax credit
2. Energy efficiency in commercial buildings, in which the credits available to developers have increased by 177% per square foot
3. Green hydrogen, which will be eligible for a new credit
4. Sweeping improvements to carbon capture and storage technology incentives

The bill has already had an impact, rightly or not, on clean energy stocks, with a broad selection having outperformed the MSCI ACWI by 16% since July 27, the date Senate Democrats first announced the bill. The outperformance has primarily been driven by solar, battery storage, and hydrogen, which, as noted above, appear to be potential beneficiaries.

I have six observations about the IRA viewed through the lens of energy infrastructure and capital markets that I think are important for tempering expectations about the future impact.

In summary, I believe that the estimation errors around total federal costs and greenhouse gas emission reduction projections are too wide to be useful data points. All else equal, the bill is an incredible shot of adrenaline to environmental efforts and the build-out of new energy infrastructure. However, I don't find ourselves in an "all else equal" world. I am concerned that inflationary pressures are a significant consideration to miss in a bill whose name pays homage to the problem at hand. Additionally, modeled impact assessments assume a manufacturing ramp-up from a cohort of businesses that don't generate money. From an environmental perspective, the most glaring hole in the bill is the lack of permitting and transmission reform. More broadly, the legal interpretation of the bill from an enforcement perspective will determine directionality, which, in many cases, we do not have yet and may take years for courts to resolve. Finally, the bill is likely just the first of many bills to come out of the world's governments that more deeply embed the governments’ steering hand in the energy economy. My ability to ascertain "what comes next" through market forces and price signals begins to weaken as governments take a more central role in resource allocation within the economy.

1. The estimation errors around cost and impact are enormous.

Several reputable institutions have released best-effort modeling results on the bill’s potential impact. Princeton University's Zero Lab estimates that the IRA will drive nearly $3.5 trillion in cumulative capacity investment over ten years and cut annual emissions in 2030 by an additional ~ 1 billion metric tons below current policy. The $3.5 trillion in investment suggests that the $370 billion earmarked for the bill over ten years would equal roughly 10% of the total investment. Most of the bill is reserved for tax credits and thus relies on incentive mechanisms to spur additional investment. In an August report, Goldman Sachs notes that the $3.5 trillion investment projections may be low considering that solar, which makes up a significant chunk of the incremental investment projections, is eligible for credits upwards of 30% of total investment. If much of the bill is earmarked for tax credits, and those tax credits tend to be 30% of the value of the investment, then $370 billion can be roughly translated into a low-end range of the potential subsequent investment, roughly $1.2 trillion. In short, tax credit incentives can produce a wide range of outcomes.

Pick any bill component, scratch one level below the surface, and it becomes evident that headline "cost" and "impact" figures need to be taken with more than just a grain of salt. Credits on carbon capture provide an example. Congresses' Joint Committee on Taxation believes that the carbon capture tax credits will cost $3.2 billion (according to a July 28 estimate of the budget effect of the bill). The only issue here is that Congressional estimates typically use historical growth rates to forecast expansion, and carbon capture capacity in the U.S. has seen little to no growth in the past twenty years. Bloomberg New Energy Finance, which tracks projects that have a good chance of being built in the next decade, estimates that the $3.2 billion cost may be closer to $100 billion, assuming a 4-fold increase in capture capacity by 2030. Back at the Princeton Zero Lab, their modeling suggests a potential 13-fold increase over the same time frame. Whether costs scale linearly here is almost irrelevant; a 13-fold increase will balloon initial Congressional Budget Office (CBO) cost projections.

The cost estimates from CBO are too low or the emission reduction estimates from reputable energy institutions are too high. Either way, we find little merit in propagating and relying on headline figures. This is neither good nor bad, but simply a recognition that the net effects of this bill will depend on how people and enterprises respond to incentives, what technology improvements may occur, etc. It's not sensible to expect or rely on too much precision at this juncture.

2. Inflationary pressures seem like a significant consideration to miss in a bill whose name pays homage to the problem at hand.

The macroeconomic and geopolitical backdrop we are currently experiencing is far more volatile than what appears to be considered in IRA modeling efforts. In every analysis I have examined, the total cost of ownership for [pick your technology] is assumed to decline precipitously. This is likely to be true for technologies that have yet to reach commercial scale but may be a bold assumption for more mature technologies such as wind and solar. Over the last year, costs across the clean energy complex have risen far more than 30%. Transitory? Perhaps. But it seems that supply considerations are now in the driver's seat regarding costs, not demand. The decline in solar power costs over the last decade is well documented and often attributed to the "power of the learning curve," or, more broadly, Wright's Law. What is never mentioned is that polysilicon costs per megawatt collapsed over this period as capacity additions from China flooded the market. It's not evident that "learning curves" have anything to do with it. Numerous questions about the economics of future supply chains and their impact on costs were not asked and answered in either the modeling of this bill or by its authors:

• Hypothetically, what happens to solar cost if 100% of polysilicon must be outsourced from Western suppliers for geopolitical reasons? The cost increases we have seen this year would likely pale in comparison.
• What happens to the cost of polysilicon if the energy used to create the product must be sourced from renewable resources? Costs go up, not down.

In July, public documents show that NextEra requested an increase in the contract price with California utility PG&E for a significant battery storage facility to ensure "the project remains viable." Among the cited issues include commodity price increases, supply chain constraints, labor cost concerns, and a rise in the cost of capital. These are all issues this bill does not address, except tangentially through government largess, and which it seems reasonable to expect will continue. The project will not be built without a material change in how PG&E compensates NextEra.

The IRA has the potential to deliver enormous value to several industries. That value proposition may be significantly muted in a world of rising costs and volatile geopolitical relationships.

3. Even if we assume a negligible difference between real and nominal dollars and that globalization continues unbounded (i.e., economic efficiency is optimized at the cost of resiliency in the face of relatively calm geopolitics), the projected ramp-up in manufacturing will be challenging.

The Princeton Net Zero lab projects that the bill will incentivize annual additions of wind and solar to grow from 15 gigawatts (GW) and 10 GW, respectively, to 39 GW and 49 GW by 2025 — less than three years from now. That’s roughly a 2x increase in wind and5x increase in utility-scale solar in annual capacity additions.

This places an enormous strain on manufacturers, and it's not clear this can be accomplished in the timeframe suggested, especially not profitability. The median profit margin for solar manufacturers over the last ten years is -1.4%. In the previous five years, it is -1.6%. Wind manufacturers are in the same boat; a medium profit margin of 0.6% over ten years, -0.24% over five years. The companies considered and the periods examined are important but don't change the story.

These companies struggle to generate a profit, and the unit economics of the business are incredibly tight. Additionally, in the U.S., there are no companies whose manufacturing utilization is less than 75%, suggesting that a national build-out to accommodate a doubling of annual capacity in less than three years is necessary. Accomplishing this at a time of rising labor costs and tightening credit conditions will be challenging. The alternative is to lean more heavily on growing Chinese production; China may also not have enough capacity in this period to serve all the world's demands.

I suspect this build-out is forthcoming, albeit on a prolonged timeline, and will require significantly more legislative assistance to ensure the requisite capital is on hand to sell products at a loss. The alternative here is a pass-along of electricity and energy costs to customers on order of magnitude that is likely not acceptable to any public utility commission in this country.

4. The lack of permitting reform and transmission investment/incentives is concerning.

A significant gap in the IRA is the lack of permitting process reform for energy infrastructure in the U.S. There is also little in the way of support for transmission infrastructure, which was specifically put aside, as permitting is considered the most significant impediment to efficient transmission investment and build-out.

Modeling efforts of the IRA appropriately note that results may vary greatly depending on complex constraints, such as the pace and scale of permitting, the expansion of transmission infrastructure, and the ability to hire and train a labor force. In many regards, simply noting that outcomes may vary fails to capture the risk that a lack of permitting reform and transmission investment poses to the potential of the IRA.

I cannot emphasize enough that absent permitting reform and a serious effort to increase transmission infrastructure, the meat of the IRA bill may be toothless.

Senator Chuck Schumer is looking to bring a vote to the Senate this fall on permitting.

5. Legal interpretations and fine print will matter.

Incentives are hard to model. They become even harder to model when language has yet to be litigated or officially interpreted. There are sections of the IRA for which we do not know which government entity is responsible for implementation.

The IRA provides tax credits for purchases of electric vehicles (EVs). To qualify for recognition, EVs sold in the U.S. must be assembled in North America. By 2024, eligibility standards increase to restricting the battery components to those not made or assembled "by a foreign entity of concern," which may include much, if not all, of Chinese manufacturing. A year later, those batteries must exclude "critical minerals" extracted, processed, or recycled from the same foreign entities of concern. If the Treasury Department interprets this law strictly as written, that would disqualify every vehicle available today. It seems reasonable to conclude that most vehicles will not meet this high standard, even if the U.S. takes enormous strides to invest in domestic extraction and manufacturing. That would likely require significant permitting reform. As with much of the bill, the effect of EV credits will be partly determined by how the Treasury interprets the law.

The White House has also picked up additional responsibilities in the IRA. Under the broad umbrella of Clean Energy Financing within the bill, Department of Energy loans must now pass through the White House for approval. There is plenty of precedent for U.S. presidents to delegate this type of authority to a cabinet member, but we have no details yet for how this will play out. It is an inter-agency hurdle that is likely to be ironed out, but it also puts plenty of power in the hands of the president (whoever that may be) to approve/deny loan guarantees at a whim.

Lastly, the IRA has amended the Clean Air Act to define carbon dioxide produced by burning fossil fuels as an air pollutant. In theory, this gives the Environmental Protection Agency (EPA) the authority to regulate greenhouse gas emissions, an authority that was recently legally challenged and struck down by the U.S. Supreme Court in West Virginia vs. EPA.

6. Increased Government Involvement in the Energy Economy.

The laws enacted by the IRA embrace an approach to governing an economy via heavy-handed industrial policy. As Robison Meyer notes in a recent article published in The Atlantic: "the era of passive, hands-off government is over." The U.S. has a rich history of debating and legislating the role of government in economic policy. While I have my own opinions on this, I don't seek to debate this issue here, only to note that this is the direction in which we appear to be heading. This prescriptive approach will have consequences that are extremely difficult to judge based on market forces alone.

The energy crisis permeating much of the world today will be managed, to varying degrees, by governments that take over and choose to subsidize, ration, and otherwise perform the role of arbiter of resource allocation. Markets are one form of allocation, attempting to optimize between producers and consumers and transmitting information through prices. The price of energy in Europe today has ceased to be a useful signal. As economist Duncan Weldon points out: "you could double prices at this point, and we would learn nothing new." Similarly, our observation that renewable energy manufacturers may be challenged to ramp considering poor unit economics and tightening financing costs may be a moot point. Governments may just take over as primary allocators of resources. Good, bad, or indifferent, we will likely see increased government intervention in energy for the foreseeable future, laying the foundation for more dispersion across inflation components in the economy.