Global Warming and Model Risk

Jim responds with thoughts about how to think of non-GDP losses from Global Warming. You should read it. Several very smart people have brought up Indur Goklany to me now, so I’ll not comment till I become very familar with his work. One more thing about climate change:

In economics there is something called Knightian Uncertainty. In quant circles, it can be called “model risk.” In everyday circles, it can be called “you don’t know what the f*** you are talking about.” Depending on your line of work, you’ve probably been there. You see someone present a model, a presentation, a research idea, or a business investment plan, and there are all kinds of charts and diagrams and numbers and powerpoint. During the Q and A, if you are lucky, someone will say “What if you are wrong?”, and they’ll respond “well if the distribution is misspecified…” and hopefully they’ll be cut off “no, what if everything you have done is completely wrong. Where would that leave us?”

As someone doing financial engineering, it probably would have been helpful to have been asked that question more in the past decade. So I want to ask, “what if all these climate models are radically under-predicting black swams and other tail risk?” To get a sense, I used an idea from Weitzman’s paper on uncertainity and went to the IPPC “The Physical Science Basis” (chapter 10, box 10.2), and got a list of a dozen and a half models that have tried to predict the increase in temperatures. These are all peer-reviewed, and (at the time in 2007) considered the latest and best research from all the fields. How do they compare to each other:

I’m particularly interested in the second diagram. Note at the 50% likely event (.5 cumulative probability, on the Y-axis), most of them are bunched up at the 3 C (5.4 F) mark, with a few less than. On average, these models all predict the same thing. Now look at the .9 mark. This is the 10% unlikely to happen confidence interval. If we under-predicted tail risk, if there are speeding and acceleration mechanisms we did not anticipate, we’ll end up out here. And here the models are all over the place. You can pick any degree between 4 C (7.2 F) and 8.5 C (15.3 F)* and find a model to support it. There’s a bit of a mass around 5 C (9 F) increase, but not like on the average.

Now that is the 10% interval. Weitzman, when he crunches this chart (and another set of companion charts in Chapter 9, Table 9.3), finds the 5% interval at, on average, 7 C (12.6 F) and 1% interval at 10 C (18 F). I won’t go further into his implications of this for pricing global warming (see here for a good overview).

I find it very helpful in modeling, especially with tail risk, to use different models, implementations and assumptions carried out from different people and see how they relate to each other. Looking at this, it seems everyone is in agreement on average. But if things go worse, it can go way worse than expected. Now that we’ve just lived through an empirical experiment in how well the best modeling can predict tail risk, I tend to look closely at that 10% marker. And the uncertainty there has me worried.

I’m reading this as “on average” everyone is in agreement, but “if things go worse than planned” everything is up for grabs, presumably because everyone is looking at different things that could go wrong.

* – See what I meant by having the Fahrenheit unit there?**
** – I’ve started Infinite Summer, so get ready for lots blogging footnotes.

Framing, The Metric System, Audience

I’m no scientist, and I’m no expert at climate change. I’m trying to learn a lot about global warming on the fly so I can be an informed citizen when it comes to what it currently going on with the debate. I kind of hate blogging on it, because I don’t like blogging on things where I’m not fairly certain I know fully what I’m talking about, and there are so many dimensions from climate modeling to political economy that it is kind of overwhelming.

So can I be completely honest? I didn’t realize, until I was halfway through trying to derive the Nordhaus Model, that the units weren’t Fahrenheit. Everyone, including myself, keeps saying “we expect there to be, on average, a 3 degree increase in temperature by 2080″, and I always implicitly mentally mapped that to Fahrenheit and not to Celsius. Be honest – did you? To scientists, and noted Francophiles like Jim Manzi, it is very natural to use the metric system. But to everyday Americans, that’s really an increase of “5.4 degrees (F)” in how we discuss weather.

(Feel free to mock the provincial Midwestern Mike if you want, with his natural discomfort with the metric system. But I’ve just been informed by a coastal elite that “Middle Americans” make the most reckless and passionate lovers, so I’ve got that going for me.)

Here’s an interesting question – should journalists report, and bloggers discuss, this to American audiences in Fahrenheit? Celsius? Both? When I say “It’s going to be 70 out” or “it’s going to be 5 degrees warmer than we expected” to my co-workers or family, it’s very clear what units I’m talking about. It seems trivial, but when a newspaper article says “An average increase of 3 degrees (C) over 80 years” or “An average increase of 5.4 degrees (F) over 80 years” it has a very different cognitive response from me, even though they are the same statement. The 5.4 seems like a bigger deal, and it is in the right units for how I think of weather day in and day out.

Buy American! Use Fahrenheit!

I know I’m forever kicked out of the respectable scientist community for suggesting the media discuss this in non-metric units. But I wonder how much “political capital” is being left on the table by phrasing it in polls and the public sphere in units not natural to regular discussion for Americans? Should the environmental blogs should start using (F), or dual-units, when writing and discussing this for proper effect? This might be a minor effect, but F scales faster (1.8x) than C, so a modest increase in C is a bigger increase in F, and the cognitive bias is for those who are used to talking in F to underestimate converting the increase in C. To say “the tail risk in an increase of 7 degrees” leaving the C out, versus “the tail risk is an increase of 12.6 degrees (F)” would probably make a big difference when it comes to “how much would you spend” questionnaires and other related public sphere discussions.

What was a subprime loan modeled on?

Even after the CJR demolished it, John Carney lays out his full theory on how the CRA caused so many subprime loans here. I think this is an important point to clarify, as it’s going to float out there in the popular consciousness. It’s a good question why so many people gave out so many questionable loans, and it is natural for people to think someone made them; I think Megan concedes the argument but still wants to see the CRA dismantled. I think the CRA is worth defending, so let’s look at the argument closer.

Carney acknowledges two key points: (1) CRA loans were very profitable for the banks in question and (2) most subprime loans were from places with no CRA coverage. He thinks that CRA loans looked so profitable that subprime lenders wanted in on that, so they duplicated their efforts but to no avail.

A subprime loan is not a slightly worse CRA loan.

Subprime loans aren’t like CRA loans. Carney focuses a lot on the LTV numbers, but that is only one characteristic of a subprime loan. My favorite chart of the subprime data:

80% of the subprime mortgages expired in 30 months; they perpetually had to be refinanced. 75%+ of subprime mortages had a prepayment penalty. This is not at all what CRA loans looked like. CRA rooted for solid, longer-term mortgages. If they ever rooted for a lot of prepayment penalties and fees to get tacked onto their loans, I’ve never seen it.

Another important statistic – in Massachusetts 60% of subprime defaults were originated in prime mortgages. So a large chunk of subprime loans were really prime loans that were collapsing. Either the breadwinners were experiencing “income volatility” or their spending was out of control or whatever. Capturing the disintegrating middle-class on terrible terms is not an objective of the CRA.

I’m going to go into some new research about a favorite topic around here, the roles the consistent refinancing, prepayment penalties and fees did to change the mortgage market, and how a consumer’s bill of rights that took us back to 1982 would be a great move. In case you don’t trust a pseudonymous blogger with a free wordpress account, it’s where the elite research is going to converge when discussing this in my humble opinion. Here’s Did Prepayments Sustain the Subprime Market? by Bhardwaj and Sengupta from the St. Louis Fed:

Using loan-level data on subprime mortgages, we present evidence on the uniqueness of subprime mortgage design. We show that the viability of such products was predicated on the appreciation of house prices. In a regime of rising house prices, a borrowers avoided default by prepaying the loan…Gorton (2008) argues that lenders designed subprime mortgages as bridge-financing to the borrower over short horizons for mutual benefit from house price appreciation…Subprime mortgages were meant to be rolled over and each time the horizon deliberately kept short to limit the lenderís exposure to high-risk borrowers.

The rationality is nothing like that of a CRA loan. It was something new, something about consistent refinancing with a huge amount of fees and penalties, using jumps in the interest rate to force prepayments. They were bad-faith loans, loans that were not meant to be repaid back, unlike a CRA loan.

“Wait, Mike. I’m getting a weird sense of déjà vu. I feel like I’ve heard this before.”

Really, where?

“A loan that wasn’t really meant to be paid off. but instead to be paid off enough with high interest rates with higher jumps to force additional payments to occur, and with a lot of the value coming from fees and penalties.”

That does sound familiar.

“Hey wait – that’s how credit cards work!”

Good metaphor. Indeed, the logic of a subprime loan looks disturbingly like the logic of a credit card. If we had to backwards out what motivated someone to go ahead and try to make these subprime loans, you’d have to say they looked at the profitable credit card market and said “ya know what, let’s design a mortgage that looks exactly like that.” The credit card model is about as far away as you can get from the CRA model – and it is very easy to imagine subprime lenders licking their lips at the sweet profits the credit card companies were making moreso than the tiny CRA market.

Update: Barry just placed a $100,000 bet saying someone could not convince a jury over him about this issue. God bless the internet.

Non-GDP Costs from Global Warming

 
So I’m trying to think through a simple question: I get that acting on global warming will not pass a cost-benefit analysis when it comes to GDP numbers. But what about everything else that matters in life? How will that be impacted? Let me tell you why I’m asking (do you need a reason!?), and tell you another set of answers I’ve found.

GDP, Markets, Religion

First off, we should just go ahead and say this: most things are going to fail most cost-benefit analysis when it comes to GDP. Look at Manzi’s cost-benefit analysis again. Assuming a GDP loss of 3%, Waxman-Markey fails 10:1. Which is to say, if the expected loss to the United States from global warming was 25% of GDP, Waxman-Markey would still fail a cost-benefit analysis when it comes to GDP. You would have to not vote for it. Perhaps that makes you dislike W-M even more, perhaps it makes you think something is screwy with using the rationality of the method.

Now on some level I don’t understand why we are doing a cost-benefit analysis at the level of the country at the level of 100 years; we have an externality, we tax/regulate it, and the relevant cost-benefit analysis is done at the contingent firm level. Where firms were previously indifferent between high and low carbon, now they make marginal choices, innovate and research which has spill-over effects, and these things lead to a virtuous cycle. Markets work.

Second, I don’t know why we are just looking at GDP. Actually I do – we do because technocrats can measure it. I’m not an anti-growth lefty, but I think GDP measurements are useful for somethings, but not necessarily useful for all things, especially quality of life in a post-industrial information technology society of the 22nd century. (I’m blogging for free, instead of working a second job, for instance.) There are people out there who buy insurance, who leave inheritances for their children. There are people out there who think that God gave us dominion over the Earth is a positive obligation to take considerations other than the Fortune 500 into account. As a lapsed white-ethnic Catholic there’s nothing quite like a good St. Peter at the Gate joke, and here’s one:

St. Peter: “So, this interview isn’t going well. Let me ask you about fighting Global Warming.”
Me, passed away: “Oh yeah, I was against all kinds of measures to fight it.”
St. Peter: “Yeah, what’s up with that? The Big Man was a fan of that Earth he gave you.”
Me: “Sure. But I was really worried doing something would upset computer models of how well the NASDAQ was projected to return in the mid-23rd century, and somethings are too important.”

In the future we’ll all think like Oil Lobbyists

And I think this the NASDAQ point is important. One thing I don’t like about the GDP models, and this is one reason I’ve avoid the modeling debates, is that it causes us to de facto argue from the position of oil lobbyists. Let’s look at GDP by country:

And use wealth as a proxy for GDP inside the United States there:

When you look at the second graph, of course the costs to GDP will outweigh the benefits. However realize that everyone in both those graphs needs an equal amount of water to survive the week. We take our ‘values-neutral’ toolbox of convex optimization under uncertainty out of the toolshed, and tell it go ahead and makes sure this distribution of GDP doesn’t get upset no matter how many people go thirsty. It causes us to reify a column in an excel spreadsheet, an indicator of how the investments of that top 10% are doing, as having more agency and claims as our actually lived lives.

Now of course you can saw that markets may have gotten us into this mess, but they’ll optimally lead us out. I hear that with arguments that globalization, water markets, transnational and continental migration, etc. will take care of any of the problems associated with global warming. To me, that requires a level of belief in how well End of History global markets will work that makes Francis Fukuyama look like a white girl with dreadlocks protesting the WTO, so it is not for me at this moment.

In the future, we’ll just camp more.

So what can we use? I’ve been flipping through Warming the World, the modeling overview of the Nordhaus RICE-99/DICE-99 models used for these debates, for clues as to how to rationalize that whole non-GDP portion of our lives under global warming. Now Nordhaus is an economist doing an economic analysis and shouldn’t necessarily be held accountable for these things, and he does support a carbon tax; however since the cost-benefit analysis from his models is leading the debate, I hoped he had an answer.

We get a few. Migration and the movement of cities and people get a GDP cost. Health issues are examined using projections of what we know of previous temperatures (and not diseases from the mass ecology destruction and migration and conflict); an average American will lose .5 years off his life, an average sub-Saharan African will lose 11 years (!!!) off his life. But here’s my favorite: All non-GDP time is leisure time, right? So (assume) anything that isn’t GDP can be accounted for by leisure. So why not just ask people what they do for fun, in the form of time-diaries, and see if global warming will impact it? That’s what they do here, in what is the major non-GDP portion of their modeling and estimates:

I kind of want to bury this answer for my great-grandchildren, so when they wonder why we stood by and let the Earth heat up they can know it was the result of us thinking camping is more fun than skiing, and we’d get to camp more on average if the Earth heated up a few degrees. So 22nd century, feel free to thank us.

I’m probably more with Manzi’s camp than I’d like to admit, though I seriously don’t know how to hurdle these objections. I really like this entry by Matt Steinglass listing some thing not covered by GDP measurements.

Cost-Benefit, Non-GDP, Global Warming

I should have been a much better lefty quant geek when it came to following global warming tail-risk models. I apologize internets, but there are only so many hours in a day. I do want to point out a few quick things from vacation with everyone talking about Waxman-Markey these days.

Let’s assume the world temperature is raised somewhere between 2.8 and 3 degrees during the 21st century. Now thanks to an argument of a fantastic cost benefit analysis by Jim Manzi and The American Scene, the USA will experience no GDP loss during this period as a result of Global Warming. Our factories will make as many widgets as they would have before, the malls will buy and sell the same volume of goods. I agree that is what our current models tell us. Let’s talk about what will change non-GDP wise.

Water, Food, Biodiversity

Jim gets this analysis on GDP from page 802, Chapter 20, Chart A on the IPCC. Now that chapter – “Perspectives on climate change and sustainability” – is mostly about ecology, with some GDP thrown from other sources. Since Jim has opened the door to this chapter, let’s reproduce some other charts and data from that chapter of what a no GDP loss world looks like.

Note that under any scenario, no North American will go hungry as a result of global warming. Under the various scenarios (use A2 and B2) around 150 million Africans will go hungry. An additional 150million people in other parts of the world. Now maybe high C02 levels will cause a boom in crop yields, and we’ll have more food than before (the parenthesis numbers represents a tail lower bound on high positive impact on C02 on plants). So there’s a chance we’ll have little impact there. That’s a rough gamble to take. Note the high numbers for water risk. We may be able to visualize that better in the next graph, though note that it is everywhere.

Let’s check out how water and the rest of the planet does. Here’s a much more interesting chart from the same IPCC chapter (I encourage you to click through for a closer look):

This is impact across zones by degree. Look at the 3 degree mark (bottom axis). We experience no GDP loss at 3 degrees warming. Something like 1.2billion people have “increased water stress.” What a delightfully technocratic term – it means that hundreds of millions of desperate people will wake up wondering “how do I get enough water this week to survive?” (I’ve heard that’s a terrible way to go.) Mind you, this newfound daily stress won’t net impact negatively how efficiently our factories make widgets or how much is sold or bought at the mall. 25% of species will be at increased risk of extinction, but our GDP numbers won’t slip.

That said, is it in our financial interest to try and reduce it? I’d like to say we don’t really know the impact of what mass migration of desperate people seeking water, or very, very rapid changes in mass ecosystems will be on our lives, and even GDP. Like a nationwide housing crash, we simply have no data to extrapolate from. I would be very surprised if they biased upwards.

Planet Earth as Sunk Cost

That said, the thing that worries me about the Cost-Benefit Analysis (CBA) approach to this issue is that there is always an implicit “do-over” cost in CBA. If we start a marketing campaign, hire a new team, or build a factory we do a CBA. If it turns out wrong, we simply stop the campaign, lay off the new team, or dynamite the factory. We eat our sunk costs and are back at square one. Sometimes this is costless, sometimes you have to pay for the dynamite. Indeed estimating the cost of this metaphoric dynamite should be high on the list of the CBA.

Now if 100 years from now, we want to “do-over”, how much will it cost to ‘dynamite’ the previous 100 years of warming? How much of GDP will we have to spend to get back an additional 10-20% of biodiversity? I’m worried that we are looking at Planet Earth as the sunk costs in these CBAs, and that makes me very worried, and being very worried makes me more willing to spend. We can’t just jump to another Earth if we got it wrong, in the same way we can build a new factory if our projections were off. This isn’t even tail risk or ‘precautionary principle’ land – we, as managers of Firm Earth doing a CBA, want to know the costs of getting rid of that 3 degrees of warming. As far as I can tell, they will be very, very high. With that in mind, seeing Conor Clarke graph out a cost of an insurance policy to build against this doesn’t seem so bad, and that positioning the numbers from another direction (a fantastic argument through that link) makes it seem appealing as a firm manager.

Update: I continue my thoughts on this issue in the next entry.

CRA Again?

John Carney has posted three ways the CRA caused lax lending. Reminds me of last fall. Barry and Felix have already responded; I want to address Carney’s points more directly:

Barry Ritholtz has been a prominent critic of the theory that the CRA has some culpability for lax lending. He has pointed out that 50% of subprime loans were made by mortgage service companies not subject comprehensive federal supervision…

2. The Threat Of Regulation Is Often As Good As Regulation. It is highly misleading to claim that just because mortgage companies were not technically under the CRA that they were not required by regulators to meet similar tests. In fact, regulators threatened that if the mortgage companies didn’t step up to the plate by relaxing lending standards they would be brought under the CRA umbrella and required to do so.

I don’t follow this. Non-depository institutions were worried about being brought under the influence of the CRA, who wanted them to make subprime loans, so they made a concession to the CRA of….creating the large majority of subprime loans?

I can easily imagine a story where the CRA forces Bank A to issue 100 subprime loans. NonBank B is not subject to the CRA so they don’t have to issue any, though the CRA is threatening to bring them under their regulation stick. So NonBank B issues 20 or 30 or 80 subprime loans to keep them off their back. But I can’t see a scenario in which they go ahead and issue 400+ subprime loans to keep the CRA off their back. (80% of subprime loans were from lenders with no regular government supervision; 50% with none at all.)

3. The CRA Distorted the Mortgage Market. With banks offering mortgages with high loan to value, delayed payment schedules and other enticing features, the mortgage companies would have quickly found themselves unable to compete if they didn’t offer similar loans.

I think this is actually a big deal – within the subprime shadow banks. Shadow bank A includes no-docs, so in order to keep their supply of mortgages to off-load up, Shadow bank B has to do the same. The worst kind of Nash equilibrium.

If Carney’s direction holds in this argument – that regulated banks took the lead in this and the shadow banks followed suit, it is a very simple quantitative argument. The data should show that Depository institutions started issuing subprime mortages earlier, and always in greater numbers, than non-depository institutions. That’s the equilibrium. The shadow banks have to “keep up” in this narrative, so they shouldn’t ever be ahead or earlier in subprime origination than depository institutions. I’ve never seen a data set that pass this hurdle. Lots of smart quants have looked at this and found the time series of originations doesn’t hold up.

1. The Creation Of Artificial Demand For Low-Income Mortgages. Banks that were regulated by the CRA often found it difficult to meet their obligations under the CRA directly.

Another narrative that is necessary is that the CRA had to be gaining, or holding steady, in regulatory power under the Bush administration. Especially during 2004-2006. I’ve been really fascinated lately with just how terrible the OTS was in its regulatory responsibility, and am writing this mostly as a way to document the following. Instead of decades old pamphlets or old stockholder speeches, let’s look at Federal Register/Vol. 70, No. 40/Wednesday, March 2, 2005/Rules and Regulations:

DEPARTMENT OF THE TREASURY
Office of Thrift Supervision
12 CFR Part 563e

ACTION: Final rule.
SUMMARY: In this final rule, OTS is making changes to its Community Reinvestment Act (CRA) regulations to reduce burden, provide greater flexibility to meet community needs, and restore the focus of CRA to lending.
Specifically, OTS is providing additional flexibility to each savings association evaluated under the large retail institution test to determine the combination of lending, investment, and service it will use to meet the credit needs of the local communities in which it is chartered, consistent with safe and sound operations.

Sounds like in the middle of the subprime mortgage market the OTS cut the balls off the CRA when it came to big lender institutions in the name of “additional flexibility.” However this might be lawyer trickery in the language. Let’s look at how the lobbyists lined up in this action – continuing:

IV. The Comments
A. Overview
OTS received approximately 4,200 comments. The vast majority (about 4,000) came from consumer and community organizations and representatives (Consumer Comments). These included community development advocates, Community Development Corporations, Community Development Financial Institutions, housing authorities, consumer protection and civil rights organizations, faith-based organizations, and educators…These comments opposed the proposal…

In contrast, OTS received a couple of hundred comments from financial institutions and industry trade associations (Financial Institution Comments). Almost all of these supported the proposal, including the portion on assigned ratings…OTS considers the level of support significant.

So the OTS finished doing an overhaul of CRA in 2005 that community organizers protested and large banks applauded. This overhaul had probably been de facto law since Bush showed up, and in 2005 there was the talk of “The Permanent Republican Majority” so it would have likely stayed. That doesn’t sound like the CRA was something to fear during the Bush years – OTS was the best site for regulatory capture (that’s why AIGFP chose it as their regulator), and they were definitely in the business of making banks CRA free.

I think it is important to remember that the characteristics of subprime loans – high prepayment penalties, consistent and high turnover to the tune of 80% expiry withinin 30 months – are not consistent with the types of loans the CRA was rooting for.

Chicago Vacation

And on that note, I’ll be in Chicago for the next week on vacation.

Hot Doug’s

Some people never got over an old girlfriend. I never got over getting to eat at Hot Doug’s every week when I lived in Logan Square. “Bacon and Cheddar Elk Sausage with Goose Island Beer Mustard and Serendipity Cheese.” “Blue Cheese Pork Sausage with White Peach Puree, Rum-Infused Dried Fruit and Roasted Almonds.” $8.

I’ve tried moving on. People tell me there are other gourmet sausage stands I should try and meet. A friend set me up on a blind date with Rosamunde Sausage Grill in San Francisco’s Haight. I got there, and was willing to give it my best. I ordered a “Duck and Fig” sausage and went to order fries. I didn’t expect there to be fries cooked in duck-fat, like Doug used to make for me, so I didn’t want to be judgemental.

They told me that they didn’t have fries, but I could get a bag of potato chips. I had to leave, realizing there would be no moving on.

Best Prank Ever – Steve Albini

So I’ll get Hot Dougs this week. Also I’ll get to see a friend who is one of the people responsible for this prank. I can’t embed non-youtube video on this free account, which is a shame, so all I can do is highly encourage you to watch that link. It’s a 3 minute description, by Steve Albini, Chicago musician, icon and the producer who brought you the Jesus Lizard, “Surfa Rosa”, the last Nirvana album, etc., describing the best prank ever pulled on him.

It’s better when he tells it in that video, but if you hate clicking through: Shortly after Jerry Garcia died, there was a giant hippie gathering in Grant Park. My friend and his friends put up signs around Grant Park saying “call this number and leave a message saying how much Jerry Garcia meant to you and we’ll make sure it gets to Jerry’s family.” The number they put on it was Steve Albini’s. If you aren’t aware, Steve Albini’s musical style does not go well with the Grateful Dead. The flyers got photocopied and passed around deadhead communities – so he received phone calls for months of hippies crying about Jerry Garcia. Best prank ever.

Added bonus: It’s not mentioned in that video but there was a second number on that flyer – The Baffler’s office. I can’t describe to you how happy it makes me to think of Thomas Frank, who is fleshing out “The Conquest of Cool” in the mid 1990s, showing up to the office and having 50 messages from deadheads crying about Jerry Garcia on the answering machine. Wonderful.

Innovative Knock-Ins

I know what the old time readers are thinking: “Mike, this blog used to be indecipherable to all but 4 people. Now you are seriously going to ignore trying to value that knock-in option on the reverse convertible bond and posting distribution charts, and instead talk about policy recommendations? Sellout. Why don’t you just call this site the ‘Radio Friendly Finance-Opinion Unit Shifter’ blog?”

*sigh*. I know. Shameful. James caught this: “(The terms can depend on whether the stock ever went below the threshold and where it is at the end of the period, which makes the deal worse for the investor, but that’s the basic idea.)” But how much worse? Let’s go under the hood of this embedded option-within-an-option so I can show you what financial innovation looks like. It gets a little intense under the fold.

Consumer Protection: Reverse Convertibles

Since we’ve already run new stress tests out of this webpage, I’m going to go ahead and run the Consumer Protection Agency out of the Rortybomb Blog while President Obama and team continue to get the legislation worked out. So welcome to the RCFPA – Rortybomb’s Consumer Financial Protection Agency. Today we also have the responsibility of the Financial Product Safety Commission. Let’s see what the thought process of such a government agency might look like. Our principles for consumers interacting with financial innovation:

1 – Markets Work; No Free Lunch
We trust that prices that undergo more market interactions reflect more information. If you are earning extra money on your investments, barring very clear reasons why markets have failed, it is because you are taking on more risk.

2 – Reduce Transaction Costs
Transaction Costs are wasteful and destructive to capital, and should always be avoided whenever possible.

3 – People are Poor with Estimating Tail Risk; They Underestimate it
People should be discouraged, though not prevented, from profiting by taking on excessive tail risk. If they do want it, they should be made very aware of what they are doing.

4 – People are Poor at Understanding Embedded Options
When embedded options are hidden inside financial products, they should be made very clear, and if reasonable, broken out into another product.

That’s a good start. Following these guidelines for consumer interaction with financial innovation will make the financial system stronger, more equitable and more stable. And I don’t find any of those four controversial.

So our first case at the RCFPA comes from Felix Salmon – Reverse Convertible Bonds, which the WSJ just called a nest-egg slasher. This bounced around the blogosphere yesterday. Here’s James Kwak on the instrument:

In a reverse convertible, you give $100 to a bank for some period, like a year; it pays you a relatively high rate of interest, say 10%. The $100 is virtually invested (no one actually has to buy the stock) in some underlying stock, like Apple. If at the end of the period the stock is above a threshold, like $80, you get your $100 back; if it is below the threshold, you get the stock instead. (The terms can depend on whether the stock ever went below the threshold and where it is at the end of the period, which makes the deal worse for the investor, but that’s the basic idea.)

The Wall Street Journal article follows the story of a radiologist who buys the instrument expecting to get a safe bond plus a free stock if things go bad and is surprised that he was written out of the money put options on stocks.

Our verdict at the RCFPA: FAIL. This instrument should not be allowed to retail investors, or perhaps not at all.

Hey! I’m a radiologist who wants to buy this instruments. What makes you want to nanny-state over me?

First off this is really selling people an embedded put option but calling it a bond. FAIL Rule #4. I’m mad he beat me to it, and so well, but here’s Nemo nailing this out of the park in comments:

As jck points out below, this is actually a naked put. (Same risk/return as a covered call, but a simpler and more accurate description when you do not already own the stock.) Since markets are efficient — at least in this sense! — there is no possible advantage whatsoever to anybody, except the bank, who is ripping people off. If you tried to use these things in volume, you absolutely would move the stock, just as if you sold lots of puts. And if these gadgets were more cost-effective than simply selling puts, somebody would arbitrage the difference away.

Therefore, this is simply a way to convince naïve investor to sell puts, but with the extra risk of the bank as a counterparty and with the extra friction of having the bank as an intermediary.

So it also involves taking on extra transaction costs when they could be avoided by just buying a put option. Rule #2 FAIL. Indeed this strikes me as more destructive of capital than breaking it down into components.

It also is FAIL on the crucial Rule #1 – It believes that the market price of a put is somehow inefficient. This is all I can read where its defenders, such as Nick Schulz say: “They are cash-secured put options.” Asking Nick if he doesn’t believes in markets is almost akin to asking him if he doesn’t believe in America and Apple Pie, but I need to ask – does he not believe in markets? Does he, and people like Daniel Indiviglio, believe there is a Free Lunch to be had by structuring cash and a put option this way?

I think I was confusing a few months ago when I said that believing in efficient markets should cause us to doubt most financial innovation. This is exactly what I mean. I believe the best price of a put option on a stock is going to the market and getting the price of a put option. I believe that the implied price of a put option, married with odd extra risks and transaction costs, in this reverse convertible bond instrument is a worse price for the put option than the market one, because I believe the more markets vet prices the better they are. Nemo’s point is absolutely correct – if they were much different, the market would be arbitraging them away.

And if you got a better deal taking this reverse convertible bond instead of just a market put, which I highly doubt but let’s assume, is it because there was a market failure that had been innovated away, thus increasing value? Or have you simply just piled into a new risk factor that you are being compensated for? Again, if you believe in markets the only sensible thing to do is assume the later until it a very clear argument has been made for the former.

Hey! Maybe I want to take on those extra risk factors so I can be compensated for them.

Really?

Yes. What are they?

Well, as Nemo said, you’ve taken on counterparty risk. If you bought, say, a Morgan Stanley Reverse Convertible Bond on AT&T (SEC – html), if Morgan Stanley isn’t around you aren’t getting anything. If you bought a put option and secured cash, you wouldn’t have that to the same degree. In so much as that is hidden tail risk, it violates #3. If you want to bet that Morgan Stanley is going to be around in a year, and be compensated for that, there are other options available that directly do this that are traded with more information.

If you dig in that SEC document there’s a bunch of risks with stock-splits and extra dividends and accerlations if AT&T goes bankrupt. Fine print risk.

There’s also that knock-in portion, of which Felix says “pricing these things is pretty much impossible.” Here’s a screen shot from Wilmott’s intro guide of just part of the equations you’d have to solve to estimate its value:

The answer to the value doesn’t have a closed-form solution, so you’ll need to set up computers to iterate simulations to solve it. At Morgan Stanley there’s a room full of Math PhDs who instead of developing solar energy are spending their hours trying to rip off radiologists using math. Buying this instrument is a bet that you can solve those equations better than them. Good luck. In practice, it means more hidden tail risk in the form of another embedded option, so FAIL #3 and #4.

Ha! I fooled you all. I’m actually a very sophisticated investor, and Felix, James, Nemo and you were wrong to be looking at this by charting the stock price. I’m actually betting that the volatility of this stock has been overrated, and am betting it will come down.

That’s very sophisticated for a radiologist who just wanted to buy a bond.

I know. Thank you. Now take your laws off my bong, man.

Hey, I’m not at all against you make esoteric bets with options. If you, Mr. Radiologist, wake up in a cold sweat after having dreams that the market is overestimating the term structure of the second moment of AT&T stock, go ahead and place a bet (I have those same dreams myself). Go short a straddle. Just sell the put option. There are baskets of derivatives that can handle this. The RCFPA should help you achieve your dreams.

What I do want though, as Chief of an imaginary government agency, is for you to make that statement in the positive. I want you to say clearly “I am selling a put option on AT&T” or “I am betting that the vol numbers will decrease.” I don’t want you buying what you believe is a magic bond only to realize later you’ve taken a large position in tail risk.

What should be our next case?

Regulation Week: Consumer Protection

Sigh. So much for a Justice League of financial regulators. It looks far more like the Articles Of Confederation. Felix Salmon has the scoop: ” FOSC = NBS + FDIC + NCUA + SEC + CFTC + FHFA + FOMC + CFPA + Treasury.” Can we just go ahead and add The Coast Guard and The League of Women Voters to the mix?

Let’s think positive thoughts. At least the CFPA is official! Obama is proposing a Consumer Financial Protection Agency (CFPA). I think this is a great move – I don’t know how much opposition there was to the CFPA, but it appears to be there and will have some teeth on it.

“Consumer Protection” is a vague notion; a lot of people want it though what it will do is often left blank. In the document it appears one big part of it will be consolidating the interpretation and carrying out of several mortgage related regulatory acts (HMDA, TILA, CRA, etc. see page 53).

It may rely a lot on “Nudges.” I am of two minds towards the “Nudging” effect. On one hand it is very important; people do over-and-under estimate certain types of risk. Information can be presented, and indeed is, in a way that is sub-optimal for people to understand their rights and obligations. This amounts to a transfer from those writing the contracts to those signing off on them.

On the other hand, people do not need to be “nudged” into understanding that a core basket of health care and education is being priced out of the range of working-class people or that ‘income volatility’ is making their lives more unpredictable and unstable. The thugs at the collection agencies calling them at 8 in the morning saying threatening things to their kids are nudging them plenty. Nudging can look like a neoliberal band-aid on a gushing wound to me when I am being cynical. As I might get today.

Again though, positive thoughts. I want to suggest 3 things, ranging from a simple bill to broader ideas behind it, that consumers could use protection from in the financial markets. Let’s start by using a particular punching bag around these parts: let’s look at what is wrong with prepayment penalties.

Ban Prepayment Penalties at the Federal Level

Daniel Indiviglio at the Atlantic:

In continuing with the regulation theme, I wanted to consider one way in which regulation advocates might believe the crisis could have been prevented: a more highly regulated mortgage industry. What if those evil/negligent/stupid mortgage brokers hadn’t been allowed to originate wacky subprime mortgages? What if a regulation had been there to stop it? If you believe that regulation should be put in place to prevent a subprime housing bubble in the future, then this idea probably seems attractive.

But how would the government go about telling mortgage originators how to write mortgages? They’ve already got truth-in-lending disclosures, so borrowers know what they’re getting into. You’d have to go further and have specific requirements about the basic underwriting standards.

Nope. One way to do it would be to ban prepayment penalties at the federal level. In that entry, we worked out a model where banks were able to bet on house prices rising by using prepayment penalties to capture part of the rises and by making sure the loans were bad enough, with deseperate enough people with terrible enough 2-year resets, to force refinancing. Let’s peek at one of my favorite charts of the subprime market, the percent of mortgages not alive by time after origination:

30 months out 80% of subprime loans aren’t active, most having refinanced. Of those subprime loans, 80%+ had a prepayment penalty. We work the numbers in that model; on average prepayment penalties reflected 6 months interest, so around 3% of the houses values. So if the house appreciates 10% in two years, the bank gets to exercise an option that transfers 3% of that equity straight over.

Because remember, banks do not normally have a delta (exposure) on housing values. If house prices rise on a prime mortgage the only value they get are secondary (less default, higher recovery); subprime mortgages were a way for them to tap that equity. Betting on house prices directly is not what we want banks to be doing. This is a direct law that could be passed by this agency. Let’s go one level higher.

Calling Fees Transaction Costs

What’s wrong with prepayment penalties at a higher level? In general, our banking system has moved to a system of fees. What would be good is correctly identifying what is a genuine fee from what is a transaction cost, as well as making financial agents less incentivized to making a profit through collecting fees and instead through providing a good/service. Making your profit off of fees and transaction costs takes attention away from the actual underlying, the service we want to encourage.

We need to be clear on terminology, and this agency can help. In many cases, these aren’t fees. The burrito maker charges me a fee for my burrito (It is not from the benevolence of the burrito maker…), and I love him for it. In the case of a prepayment penalty, it’s a transaction cost. The proper “fee” is the interest rate on the loan. This penalty is a transaction cost – it prevents money from going to where it needs to go, and gummy ups the market. Markets hate transaction costs.

It also creates incentives for volume over quality. And it also creates incentives for origination instead of holding. The mortgage part of the financial crisis is full of stories where it is clear the banksters in questions were less concerned with a loan getting paid off and instead with it getting paid off enough. This business model is part of the reason why. This was clearly a problem in the previous crisis, and will likely be a problem in some other way in the next. But isn’t this transaction cost really just a hidden option?

Properly identifying embedded options

Let’s go higher. When I was discussing this prepayment penalty theory with a very smart person from a hedge fund, he told me that selling people embedded options is always deviously clever because people don’t understand that they are buying them, and often don’t understand their value. I think this is key.

Take a mortgage. It is simply a risk-free rate, with an option to default (put option) and an option to prepay (call option). Most consumers, if they had 50%+ equity in their homes, probably wouldn’t realize that the bank is more concerned about them exercising the prepayment option (exposing them to nasty negative convexity) than defaulting. If the consumer has that much equity, chances are he rationally won’t mail in his keys. But most consumers expect the opposite, that banks are freaked out about them defaulting and rooting for them to make extra payments more often.

We should emphasize options that smooth risk and smooth consumption. Prime mortgage buyers have less than 2% prepayment penalties. Lord knows that they could get them in exchange for some extra money, but they rationally choose to pay a little extra to avoid having to pay a lot extra under adverse conditions. Most options, from option ARMs to Prepayment Penalties, make payments more volatile and de facto encourage gambling. Hidden embedded options that make payments more volatile, harder to predict and more of a shock should be discouraged in exchange for options that, if anything, make payments smoother, easier to predict, and are less conditioned to be most adverse when things in the economy are worst.

So that’s taking one penalty and looking at it from several different levels. What are some other things I think they could look at?

Insurance

People consistently overestimate the amount of insurance they have on their house. Surveys and studies find that they aren’t rationally optimizing marginal costs but instead simply not comfortable with properly estimating tail risks and not updating contracts in their head based on changes that occur in the world. Coming up with some better contracts that nudge with people’s expectations are good for consumers.

Underbanked Communities

I do not know enough about this subject, but I have heard enough smart finance people I trust talk about it for me to believe that there could be an effort here.

Private Student Loans

This is a terrible, nasty, crony-infested industry that needs some disinfectant. Is the CFPA up for the challenge?

I’m missing a ton of obvious ones. What are some other ideas for this CFPA to tackle?