Direct Climate Markets: the Prospects for Trading Teleconnection Risk - Overview

In 2013, I finished a Ph.D. dissertation on the prospects for launching the first direct climate markets --- markets trading the risk of major regional climate anomalies. In particular, I focus on El Niño/La Niña, a climate anomaly with serious economic consequences across the globe.

I've tried to provide most of the analysis necessary to launch El Niño/La Niña trading. Based on my work on an innovative insurance project, this dissertation details why and how to launch traded markets on El Niño/La Niña. Combining sophisticated statistical modeling with qualitative interviews, this dissertation includes:

• statistical models of El Niño/La Niña's worldwide economic impacts;
• a pricing model to anchor expectations for virtually any risk management contract on El Niño/La Niña, conditional on existing forecasts;
• a comprehensive statistical description of the lifecycle of new derivatives (research co-authored by Michael Penick an economist at the U.S. derivatives regulator, the Commodity Futures Trading Commission);
• interviews with risk management professionals at businesses facing El Niño/La Niña risk and the financial firms interested in trading that risk.

I conclude that catastrophe bonds settling on NOAA's Niño 3.4 sea surface temperatures can and likely will launch in the near future.

How can you hedge climate risk?

The markets we have today...and why they don't work for climate risk.

There are markets covering the policy response to climate change, like those on emissions permits. But those markets are only indirectly tied to the climate. (Discussion here and here.) The EU's emissions-trading system (ETS), the legally binding system underpinning the largest carbon trading system on the planet, recently provided a jarring reminder of that distinction. On April 16, 2013 the European Parliament voted against proposals that would have altered the scheduled for issuing emissions permits under the ETS. That day, the price of EU carbon permits fell 40 percent. Nothing about the long-term weather tends changed 40 percent on April 16, so those markets would have surly disappointed anyone looking to protect themselves against climate extremes.

Another possible alternative for businesses and individuals worried about extreme weather are exchange-traded weather derivative, like those that allow energy companies to hedge the risk of hot summers in major US cities. But where emissions markets are too broad for climate hedgers, weather markets are too narrow. It is difficult to see global scale changes in climate just looking at basic weather indicators (e.g. rainfall and temperature) for a handful of cities. In part because of their limited scope, weather markets have failed to attract high sustained liquidity.

How El Niño/La Niña markets will be different

El Niño/La Niña is an example of what climate scientists call teleconnections - statistical and physical links between regional oceanic/atmospheric anomalies and patterns of catastrophic weather around the world. Teleconnection bridge the gap between climate, a process that unfolds over decades or centuries, and weather as we live it.

Teleconnections are excellent candidates to be the basis of insurance or derivatives thanks to the scope and diversity of their impacts, the predictive nature of the signals they generate, and the fact that they can be represented by simple indexes published by trusted national meteorological services (NMS), such as NOAA in the United States.

This dissertation explores the idea that traded markets based on these teleconnection indexes could provide a low-cost risk management tool to firms, individuals, and institutions facing catastrophic weather risk. They would also provide consensus forecasts on weather-related catastrophe risk. El Niño/La Niña appears to be the lowest hanging fruit among the teleconnections, the one phenomenon from the group that is already a target for insurance. But if El Niño/La Niña risk markets succeed, may set a precidence for other teleconnections like the Arctic Oscillation, which I discuss in an epilogue to my dissertation.

The innovative insurance behind this dissertation

In 2011, I was part of a team that helped Caja Nuestra Gente, a large microfinance bank in Peru, purchase the world's first insurance against the El Niño climate phenomenon. (This report for the United Nations Development Program provides a nice introducion to the insurance.)

That team came from a research firm, GlobalAgRisk, led by University of Kentucky professor, Jerry Skees. GlobalAgRisk and its sponsor, the Gates Foundation, initially focused on El Niño because the phenomenon caused catastrophic flooding in a part of Peru with a particularly high incidence of poverty. (Note: Southern Ecuador was similarly impacted by terrible flooding.)

The most recent severe El Niño events, in 1982/83 and 1997/98, caused annual precipitation on the order of 40 times average. More dramatically, over two weeks in 1998, a lake, second only in size to Lake Titicaca in Peru, formed spontaneously in the desert south of Piura (a regional capital that was the center of our work). These events caused devastating personal and economic hardship. Some 200,000 people were displaced from their homes, the second largest port in the country was shut down, and the regional economy ground to a halt.

In 1998, El Niño flooding left many banks in the country's impoverished northern provinces with inadequate capital. That meant that banks could not originate the new loans needed to rebuild hurting communities, a feedback loop repeated on a global scale in the 2008 Financial Crisis. To avoid a repeat of that cycle, we designed the insurance policy to protect the banks' loan portfolio against defaults after catastrophic El Niño flooding. The insurance we drafted moved part of the bank's El Niño risk to a large international reinsurance company. Thanks to that risk protection, the bank now plans to continue lending to affected communities when credit is most urgently needed - during the recovery and reconstruction following El Niño.

Our insurance addresses a fundamental social need and is a true financial innovation. Payment is trigged solely by the sea-surface temperature (SST) rise that defines El Niño, measured by the US's National Oceanic and Atmospheric Administration (NOAA) over defined regions off the coast of Peru. Sharp rises in that temperature, particularly in the months of November through January, define the El Niño phenomenon and cause catastrophic flooding in Peru.

The policy pay exclusively based on that causal temperature signal, which precedes Peruvian flooding by months. That lag between the signal and the catastrophic rains in Peru gave us the chance to create, we believe, the first regulated insurance in history that pays before a disaster. The temperature measure that triggers payment is posted on NOAA's website in early January, so the insurance company can send its clients checks before they are affected by floods between late January and April.

Beyond insurance - the benefits of risk trading

I'm proud of the insurance we put togeather in Peru. But my experience on that project underlined some important limitations of insurance and reasons why El Niño/La Niña may be better suited to derivatives or similar financial instruments that are activly traded.

The biggest advantage of moving an El Niño/La Niña index to traded markets involves dynamic pricing of the underlying index. Currently, the sales closing date for the insurance (i.e. the date when clients have to decided if they are going to buy) is a full year before the period of coverage – meaning that a firm looking for coverage during the 2013 El Niño season will need to choose whether or not to buy by then end of January 2013.

This schedule avoids the adverse selection problems created by El Niño forecasts, which are improving incrementally every year and open up the possibility of opportunistic purchases – with the sophisticated buyers only buying coverage in years where they think an extreme El Niño is likely. Indeed, in the first year that GlobalAgRisk’s El Niño insurance was on sale, a large fishing company expressed interest in purchasing coverage, but requested additional time, beyond the original sales closing date, to make a final decision. In those critical weeks, new forecasts did come out suggesting El Niño was less likely. While it is difficult to directly link the fishing company's subsequent decision not to purchase coverage to those forecasts, the experience provided a stark reminder of the adverse selection problems inherent to insurance based on a forecast-able index. In the following year, the sales closing date was moved to January.

The lag between paying for and receiving coverage under GlobalAgRisk’s El Niño insurance increases the opportunity cost of hedging and means that the market is unlikely to attract the attention of risk managers with shorter planning horizons. These are problem that would be avoided altogether in exchange-traded markets, where prices are free to move as new forecast information becomes available.

The knock-on effect of dynamic pricing would be better public and private decisions related to this key climate phenomenon. Exchange-traded El Niño/La Niña derivatives would provide public information not just about the price of risk protection but also about the likelihood of extreme El Niño/La Niña events. Currently decision makers (particularly in Peru) have to grapple with many competing El Niño/La Niña forecasts, often built using different datasets and methodologies.

International Research Institutes for Climate and Society run by Columbia University provides a running tally of the forecasts of the best El Niño/La Niña models from academia and national meteorological services. One look at that graph makes clear the need for the definitive consensus forecast that derivatives markets would provide. Without that touchstone newspapers and politicians in the most effected countries (Peru and Australia in particular) have often leaned heavily on alarmist forecasts - creating El Niño fatigue among ordinary citizens and policy makers.

Finally, El Niño/La Niña is well suited to exchange-traded derivatives markets because it would facilitate the direct transfer of risk among a diverse collection of hedgers across the world. El Niño/La Niña affects many regions of the globe and within each high-risk region some industries benefit from extreme events (such as reinsurers who historically face fewer losses thanks to suppressed hurricane activity during extreme El Niño) while others suffer. Direct risk trades between those groups would contribute to price discovery and provide sustainable liquidity.

However, all the advantages of moving El Niño/La Niña to traded markets are predicated on liquidity - high and stable trading activity. Markets will only provide reliable forecasts if experts spend time watching for changes in prices that are not justified by the current state of the climate. Individual climate scientists and the financial firms who might hire them will only dedicate the resources to watch for these mispricings if they believe that they can make profits when they catch a temporary mispricing. That is difficult when a market is small, either because there is no one to take the other side of the trade or because the few people who would take the trade can infer from a large order that they can demand a better price.

So, more than just the idea of teleconnections markets, this dissertation looks at liquidity. Can teleconnection markets generate it? Under what conditions? I've devoted the last few years to researching these questions, putting special emphasis on El Niño/La Niña. Throughout that time, I've tried to balance my personal enthusiasm with the skepticism of an empirical researcher. I've asked questions that I believe skeptics would ask, tackled them with rigorous quantitative methods, and set to explain the results as simply as possible. My results are cautiously optimistic on the prospects for El Niño/La Niña markets in the near future.

Dissertation road map - what information do we need to start El Niño/La Niña risk markets?

Part I: What are the economic impacts of El Niño/La Niña? Why is El Niño/La Niña risk a good candidate to move to financial markets?

Estimating El Niño/La Niña's economic impacts

In the chapter 1, I introduce the dataset that I used to estimate El Niño/La Niña's economic cost. In its raw form, that dataset is substantial. But, in my opinion, it is insufficient to provide a comprehensive picture of teleconnection impacts. To address that shortcoming, I infer missing data using Bayesian regressions. The chapter explains how and why I made that inference.

Chapter 2 gets to the heart of El Niño/La Niña's economic impacts. It presents a brief overview of the El Niño/La Niña cycle and my estimates, based on the enhanced database from chapter 2, of its economic cost in vulnerable regions. The chapter shows that: * El Niño/La Niña risk is large enough in absolute terms to justify formal risk markets; * large pools of El Niño/La Niña risk offset one another in time and space, suggesting that El Niño/La Niña markets could sustain balanced, direct trading among hedgers; and * El Niño/La Niña creates a pool of economic risk that is comparable to those underlying some large futures markets today.

Pricing El Niño/La Niña risk

To conclude Part I, chapter 3 presents provides a pricing model to that covers a wide range of financial instuments (including insurance, options, and futures) to manage the risk extreme El Niño/La Niña events. This chapter includes models of how the likelihood of El Niño/La Niña events changes over time, as meteorologists release new El Niño/La Niña forecasts. This chapter provides would-be market makers in El Niño/La Niña risk with a starting point for the spreads they set and hedgers with a baseline for identifying a well-priced hedge.

Part II: In what form ((re)insurance, futures, options, swaps, etc.) will El Niño/La Niña markets be most likely to reach a sustainable level of liquidity? What form offers institutions, firms, and individuals the most efficient tool for managing their risk?

What financial instuments are the best fit for El Niño/La Niña risk?

In chapter 4, I provide basic information about the forms that an El Niño/La Niña risk market could take. I also discuss my initial hypothesis that exchange-traded derivatives (futures and options) provide the most efficient and equitable avenue for managing El Niño/La Niña risk. This chapter is not original research, but an overview of what I consider the prerequisites for efficient social outcomes from teleconnections markets.

What is the baseline probability that any new cleared/exchange-traded derivative will succeed or fail?

In chapter 5, a collaborative effort with the Senior CFTC Economist, Michael Penick, I test my hypothesis about futures and options by looking at the probability that any new exchange-traded derivative will succeed. That analysis is based on a comprehensive database covering all derivatives traded on US exchanges since the mid-1950s. I show how the lifecycle of derivatives has changed over time, and what that means for innovative markets. The picture emerging from that analysis is moderately hopeful for El Niño/La Niña. The probability of reaching very great liquidity was never high, and has fallen over time. However, the probability of reaching modest levels of liquidity and the probability of recovering from years with very little trading have improved remarkably in the last decade.

What do catastrophe/weather risk professionals think about the prospects of El Niño/La Niña markets?

I finish my analysis in chapter 7 and chapter 8, by asking risk professionals, who might be early adopters of El Niño/La Niña risk management instruments, for their thoughts on the likely demand and supply of El Niño/La Niña risk management and the form that those markets should take. Chapter 7 provides qualitative analysis from more than 35 in-person interviews around the world. Many of those interviews focused on the current state of catastrophe and weather risk trading so I've include a short introduction to those markets in chapter 6.

These interviews forced me to reconsider my initial hypothesis about futures and options. While I continue to believe that futures and options remain an important end-goal for El Niño/La Niña markets, my interview subjects persuaded me that those exchange-traded markets will have a better chance at success if they evolve from catastrophe bond (CAT bond) trading.

Chapter 7 includes the proposal of one interview subject, John Seo of Fermat Capital, of a liquidity fund that will help smaller investors access customized coverage from CAT bond markets, generally reserved for institutional investors. This liquidity fund would achieve the desirable social outcomes associated with exchange-trading within the context of the CAT bond markets that industry professionals prefer. That fund would operate at no net cost to its host institution.

Chapter 8 provides quantitative validation of the interview findings. It includes an adaptive choice-based conjoint analysis of 15 experts willingness to pay for various contract designs.

What's next after El Niño/La Niña markets?

After some concluding remarks on the solid prospects for El Niño/La Niña markets the epilogue introduces the Arctic Oscillations, the teleconnection that I consider the next frontier for formal risk management after El Niño/La Niña.