Are We On the Verge of Another Financial Crisis?

Harvard Business Review - Economics & Society

PM Images/Getty Images

by Eben Harrell, December 13, 2020

Summary. In 2008, a collapse in housing prices triggered a global financial crisis. John Macomber, a senior lecturer at Harvard Business School, believes history may be about to repeat itself — this time caused by our failure to acknowledge and confront the perils posed by a changing climate. The financial system hasn’t correctly priced in the risk from fires, floods and storms. If the correction happens suddenly, the collapse in housing prices could spread through the financial system. The incoming presidential administration must take politically unpopular steps to avoid this scenario.

John Macomber, a senior lecturer in the finance unit at Harvard Business School, believes we may be on the verge of a collapse in housing prices and an ensuing financial crisis — this time caused by our failure to acknowledge and confront climate change. In a phone interview and a written email exchange, he shared his reasoning and what the incoming Biden administration can do to prevent this scenario.

You’ve been warning for years that America’s housing market has been ignoring the risk of perils associated with climate change. Do you believe we are approaching a correction?

Yes. Damage from climate change has accelerated faster than many people anticipated. In USA in 2020, there were 16 weather/climate disaster events with losses exceeding $1 billion each (some much larger). The average from 2015 to 2019 was 13.8 such events. The average for the 40 years prior to 2020 was 6.6. What’s more, we are seeing risks we didn’t foresee just a few years ago. We’ve been rightly worried about coastal flooding from sea-level rise but in the last several years there’s also been an increase in river flooding from rain and huge damage from wildfires.

Among other issues, we haven’t faced the tough question of whether people should be restricted from building or rebuilding in these places that are, in the example of California, natural fire corridors that have been recognized for centuries. Instead, in California we’ve required utilities to bring power to homes in these dangerous areas, and now the state is mandating that insurance companies renew fire policies at below-market rates. Similarly, in parts of the east coast, private insurers have long since exited the homeowner flood risk market and instead the coverage is provided with deeply subsidized premiums by state agencies relying on the National Flood Insurance Program.

This is a classic market distortion.

Indeed. It encourages people to make or maintain housing investments that are exposed to more danger than they realize. For now, governmental entities absorb the extra cost of these risks when they repair or rebuild these homes (using the tax receipts from other property owners, by the way).

Insuring, repairing, and rebuilding properties that really are uninsurable has artificially inflated home prices by papering over this risk pricing gap. In the short run many parties benefit from propping up housing prices, but with increased exposure to peril and further tightening of government budgets this cash-hemorrhaging system cannot endure. The question is whether it’s going to settle out slowly or settle out fast. My concern is that all of a sudden it just snaps and there’s this giant reset that leads to a real disruption in housing prices.

Take us through that scenario.

The optimistic scenario is that a gradual sea level rise or a slight increase in fires will lead to gradual declines (or relatively slower appreciation) in house prices. The broader system has time to adjust.

The greater worry is that insurance premium support will suddenly dry up, and at the same time mortgage underwriters will start to factor in the substantial danger of these exposures. The result will be a dramatic consequent rise in insurance premiums, coupled with a reduction in mortgage loan-to-value ratios (and at worst the complete inability to buy fire and flood insurance at all, or to refinance a mortgage). Housing prices will plummet in these areas. For many homeowners the equity in their property is their biggest asset. It’s a real problem if that asset declines in value or even goes negative (if you owe more on your house than its risk-adjusted value).

This scenario will result in a second circle of trouble. Most American municipalities get the bulk of their revenue from property taxes. Property taxes are tied to the value of homes and commercial real estate. If home values fall, then property tax receipts fall without a simultaneous reduction in a city or town’s expenses, so their ability to service their municipal bonds becomes imperiled. That could lead to the ratings of the bonds being downgraded. That puts cities and towns under cost-cutting pressure, which then leads to other stresses on government services. It also increases their cost of borrowing, with both factors leading to a downward spiral.

A knock-on effect will be a potential decline in the ratings and value of certain bonds. Tax-advantaged fixed-income instruments, such as municipal bonds, are a big part of many people’s retirement portfolios (and many insurance companies’ reserves). I argue, then, that this aspect of climate risk touches everyone’s pocketbook.

The 2008 correction in housing prices spread throughout the financial system. Is there concern this could happen again?

Yes. You won’t be surprised to hear that the really dangerous amplification is from algorithms and risk transfer in sophisticated financial products. Homeowners buy their property/casualty and fire/flood insurance policies through brand-name companies, such as Allstate or Progressive. But these companies often don’t retain all of the exposure to pay for loss events. In particular, they don’t mind being exposed if say one house burns down – the other premiums collected cover that cost.

But if an entire county or part of a state gets hit hard by a hurricane, they can’t cover losses to all of those homes on their own. They often contract, in bulk, with another tier of insurers called reinsurance companies. These firms include giant but lesser-known companies like Swiss Re, Munich Re, and General Re. Those international firms attempt to spread their exposure across the globe and across many categories of peril like tornado, hurricane, earthquake, wind, and flood.

In addition to diversifying the risks, the reinsurers also can slice off some of the risk into insurance-linked securities — including weather derivatives sometimes known as “catastrophe bonds.” The probability of an event happening and the likely cost of the event are rated by several specialty companies then bought and sold by financial investors — who have zero knowledge of or interest in your particular home or city — who can be paid to accept financial exposure of a defined nature for a fixed period of time in the event that one of the named events occurs.

This means we have a situation where whoever is buying or selling the risk is multiple steps away from the actual property. Sound familiar? It’s hard to gauge how far these instruments have spread into the financial system.

This sounds a lot like the financial weapons of mass destruction, such as the securitized instruments that were traded before the 2008 crash. What about rating agencies? Are they doing a better job of independently assessing risk?

In this instance, the entities which evaluate insurance-linked securities (and most of the reinsurance companies that trade them) all have proprietary systems for assessing various risks and exposure in the broader insurance market. Some like RMS and AIR have been modeling not only flood, earthquake, and tornado risk for years, but also perils like terrorist attacks and pandemics. Others like Jupiter and 427 focus on potential weather incidents, like wildfire, flood, sea rise, and drought.

One problem is that the inputs are not agreed. There is not consensus, at least in the United States, about existing flood risk even independent of sea-level rise, never mind about potential future rise. A second concern is that there also is not consensus about how to model what might happen. The third concern, and in my view most unnerving, is that these firms’ projections are proprietary. Modelers and the financiers that they service know more about the prospects for my property than I do. I find this information asymmetry to be worrying. Who is going to come out on the short end of the stick here?

The well-known credit rating agencies, like Moody’s, S&P, and Fitch, are behind the curve right now because they tend to focus on financial ratios, like debt service coverage and loan to value. For decades the natural disaster exposures of homes, municipal buildings, and power plants was static, and history of past loss was a very good guide to future loss. That guide is no longer reliable. That adjusted thought process will need to propagate through the industry as well, and that could change a lot of AAA bonds to BB+ and once more percolate down into collateral and swaps, as it did a decade ago — again touching many people who don’t live anywhere near the problem geographies and don’t even invest beyond their 401(k).

What can the incoming Biden administration do to address these risks?

The first step is to start talking transparently about Americans’ exposure to climate-related perils. Other than a few exceptions such as a recent report by the Federal Reserve, few government agencies have even mentioned this issue.

This means publishing a common set of projections of impact. FEMA, for instance, has sometimes been pushed to optimistically revise flood maps under political pressure. This needs to end. FEMA is making inroads with Risk Rating 2.0 which aspires to update the methodology for understanding a property’s unique flood risk based on location, nearby geologic and hydrologic features, and house construction. FEMA is also trying to allocate money to disaster prevention in addition to its classic disaster recovery — which is a much more efficient way to direct funds if done right — notably with the Building Resilient Infrastructure and Communities (BRIC) program. This is a worthy pilot, and the new administration should back it fully. However, this is not nearly enough money to reinforce all the properties that are exposed even in the short term, never mind the long term.

The second step is to start to taper down the distorting effects of mispriced flood and fire insurance. This should not happen all at once — we don’t need to manufacture a cliff when we have a lot of other concerns — but it should happen over say 10 years or so. Home owners should start making property decisions in the normal rhythm of their lives with a proper understanding of what the real exposures are. Federal and state governments should slowly exit the flood insurance business.

One hopes that the Biden team recognizes that a pricing correction is coming eventually anyway — the market and mother nature always sort out prices in the end. It will be much less traumatic if the air can be let out slowly rather than all at once.

This scenario sounds a lot like The Big Short. Are there investors currently positioning to make money off this upcoming correction?

Yes, and they are starting to take a high profile. But it’s a bit more complicated for investors this time around. It’s not clear how you actually accomplish a short trade as there’s no short market for municipal bonds or housing the way there is for stocks (the same obstacle that faced the short speculators in the mid-2000s).

I see three main ways in which investors are participating. First, for a large property owner, like Walmart or Blackstone, it’s now simple to include a short and long-term flood and fire exposure model: Which properties do you want to own, and which would you rather not own based on this criteria (in addition to all the other filters)? If they had the information, home buyers and sellers could also use this filter.

Second, it’s possible for investors to select between asset managers. For example, if you are managing your own portfolio, can you analyze three different apartment REIT stocks based on where you think there is climate change exposure? An obvious choice would be to exit parts of Florida, Arizona, and California in favor of other locations.

And finally, it may take a long time, but if in the next several decades substantial numbers of people are going to be displaced from Virginia or Alabama or Washington due to sea rise, where will they go? Maybe inland nearby — to Richmond or Orlando — or maybe to say Duluth or Rochester which both have plenty of housing, lots of infrastructure, increasingly temperate climate, and access to some of the largest bodies of fresh water in the world. Displacement on the one hand leads to development on the other.

Infrastructure spending continues to have bipartisan support. Could the Biden administration encourage infrastructure to improve resilience to climate change? I’m thinking seawalls and flood gates?

That’s a good idea. It’s a potentially bipartisan initiative. But first the country needs to agree politically about the goal. What is the objective of the infrastructure investment? Is it to improve the quality of life for a lot of Americans? Is it to increase economic competitiveness in certain parts of the country? To protect a subset of homeowners from fire and flood? Or is it just to spread money around congressional districts?

To me we should think “people first,” not “spending first.” A key project-selection criterion would be how to accomplish the most benefit for the most people, using the least resources (a concept attributed to Buckminster Fuller). Under this rubric, if it’s quality of life, we should invest first in developments that serve immediate health needs, such as water projects in Flint or Philadelphia. If it’s supporting job rebounds after Covid-19, it’s public transportation in cities.

After that, yes, certain climate adaptation tactics would pencil out economically — but not all of them. I’ve written elsewhere about investing in resilience. Interestingly, seawalls, flood gates, and dikes are popular in conversation, but they tend to be very expensive and not work very well. With respect to the housing exposure discussed above, it’s not a realistic way to defend hundreds of thousands of homes for decades. Seawalls help in a storm surge but not over the course of the long haul; the water can sneak under a seawall over time if the ground is porous. They also are only as good as the weakest link — if one property owner lets their seawall crumble, the water hits all the neighbors too. Other sea-rise mitigation projects are probably more cost effective, notably green (natural) infrastructure like berms, mangroves, and replenishing vegetation in tidal areas, as well as gray (concrete) infrastructure, like raising buildings, raising sewer drains, raising substations, and building fire breaks in fire territory.

In the end though, I expect there will be certain areas in the country (and in the world) where people just can no longer live, as a result of drought, sea rise, wildfire, or flood. This realignment has the potential to be very unfair to people with poor access to capital and poor access to information (while beneficial to those with capital and data). It also has the potential to be forward looking, well-managed, and fair. The sooner we confront this reality the less painful, and more equitable, the correction will be.

Eben Harrell is a senior editor at Harvard Business Review.