Occasionally, some Bureau Members get together and discuss economics amongst themselves. The moderators will sometimes post conversations that we believe are of high quality, for the benefit of the subreddit. Here is one such conversation. Feel free to respond to the conversation, or provide feedback on any of the ideas discussed inside.

gorbachev

Re: financial crisis

I recalling having read some JELs and JEPs back in the day that were like "so basically, the financial crisis was just a bank run"

except instead of johnny next door demanding his cash from Jimmy Stewart, who replies that it's in neighbor fred's house

it was financial institutions going to other financial institutions, who replied they don't know where it is because turns out they didn't understand their portfolios that well either

my experience w/ undergrads so far is that when asked to explain the crisis, the explanations I get are

Bad home mortgage lending decisions ----> housing market collapse ---> a bunch of financial institutions lose their shirts on misc derivatives -----> ???? -----> unemployment

I think stressing the stuff in the ???? would be a good idea

Especially because my impression is that most serious policy makers think the ?????? is the important part

Even though huge %s of undergrads will explain that the plan to prevent a future financial crisis is to ban complicated derivatives and crack down on Freddie, Fannie, and mortgage markets in general

when really the focus of policy makers has been like, how do we prevent these situations where people freak the fuck out over counterparty risk unexpectedly ballooning, stuff like that

idk, just seems we ought to stress it

ocamlmycaml

the easy causal story is kind of hard because a lot of it is happening simultaneously

https://fred.stlouisfed.org/graph/fredgraph.png?g=dBfr

mrdannyocean

what's the best summary of that '???', gorbachev?

stock market crash produces negative wealth effects, scares the shit out of everyone, hurts consumer confidence, people stop spending. Banks stop giving loans, leading to lower business spending/investing and job cuts. the consumer piece compounds and hits the businesses again, leading to further cuts.

is that basically what we should be imagining for the steps between 'financial institutions are in a panic and lose a ton of money' and 'unemployment is now 10%'?

also at the same time, consumers get scared and lose confidence and stop spending, consumers have more debt than they can handle, banks are not lending, and all of those things lead to businesses who have to layoff employees because cash flow is significantly down

basically, is there a component of this that is consumer-demand driven?

roboczar

i don't know if that's a robust behavioral story. the movements of the stock market don't tend to drive consumer behavior

losing a job due to a drop in access to credit for firms, yes

mrdannyocean

yeah the stock market thing is probably not the most serious cause

the corporate credit crunch is likely far more impactful

I do think generally reduced consumer spending could play some role even though the credit crunch mattered more. That consumer impact could be from a few factors - sudden spike in defaults, loss of confidence, etc.

And so at the worst possible time, consumers are all the sudden pulling back from spending

gorbachev

well, my understanding from the JEPs and JELs is that the ???? is something like

some institutions lose their shirts ----> other banks look at their assets, begin worrying about which of their counter parties actually can pay up ----> big freeze up in trading and lending between financial firms ----> commercial paper & repo markets freeze up too slash experience what is essentially a bank run

by that point things are so fucked that we essentially don't have a financial system for a while, companies like walmart only got the capital to meet payroll because the fed stepped in

so that's the part of the ?????? that escalates the financial shock into a financial crisis

roboczar

yeah i agree. it was basically a wholesale heart attack in the credit markets

firms can't meet operational costs across the entire economy, and then everyone's fucked

gorbachev

then you need the part of the ?????? that turns the financial crisis into unemployment

I know less about that, but I'm guessing "no credit" goes a long way

also industry specific shocks like to home building. housing wealth shocks. whatever

ocamlmycaml

here's a busier chart with more financials

https://fred.stlouisfed.org/graph/fredgraph.png?g=dBfY

roboczar

firms basically run on short term lines of credit, not revenue

gorbachev

yeah, iirc literally walmart wouldn't have made payroll if the fed hadn't stepped in and either made them a short term loan or otherwise done something to bail out the commercial paper markets

anyway, I would mention that there are cool papers tracing the propagation of unemployment across the country following the GR

like, you can pick specific places and say "the great recession started here, in terms of unemployment"

and then you can see unemployment basically transmit across input output linkages

from place to place

roboczar

anecdotally, the company i helped found in the mid-2000s essentially started its death spiral when the bank we were getting a revolving line of credit from unilaterally converted it to a term loan, effectively strangling us to death

due to the GFC

we basically went running hat-in hand to any VC firm that would have us, and eventually fire sold when we were all out of options

that was fun

gorbachev

woof

geerussell

Shouldn't overlook the role of fraud in amping up the magnitude of the crisis. From origination to conveyance to securitization a lot of questionable-to-outright-illegal stuff had to happen to keep feeding the beast with new loans.

Also mrdannyocean, with regards to wealth effects, there was a very direct one as people were literally spending home equity with lines of credit and other financial products allowing it to directly provide purchasing power as opposed to simply "feeling" richer and spending more out of current income.

gorbachev

realistically speaking, how much did fraud really contribute to the size of the housing bubble? do we have something like

state level variation in the degree of regulatory stringency, have related to the magnitude of the housing market correction?

That seems imperfect - hmm, something else?

I'm skeptical that fraud played a large role

I'd almost imagine causality running in the opposite direction

big booming housing market that seems to be showering $ on anyone walking near it

------> fraudsters want in

roboczar

i think it would be more likely that, while fraud did happen, the state of people's debt wasn't necessarily a concern, because they had incomes. It became a problem with the state of their incomes changed due to a credit freeze brought on by ?????

geerussell

Fraudulent appraisals were key in keeping year-over-year home values on the march. Fraudulent origination helped feed the raw material of new loans into the securitization pipeline. Things like MERS and simply ignoring tax constraints in securitization provided a lot of extra bandwidth. Not to mention the conflict of interest in ratings agencies paid by the same people who create the products being rated and the abstraction of ratings from credit evaluation to models provided by the people creating the products. Etc, etc.

roboczar

so yeah, fraud was an issue, but it was dwarfed by the substantial change in circumstances for holders of consumer debt that were now unemployed

geerussell

It's not either/or though.

Not all the people even had incomes, ninja loans were a thing.

roboczar

do we have stats on how much of a thing though

like i'm skeptical it was endemic

geerussell

It doesn't have to be endemic to set the pace.

gorbachev

I'm skeptical, but I should probably also acknowledge that my skepticism of the fraud explanation is partly politically motivated

in particular, I see people w/ the fraud explanation as like

delivering cover for bad policy

like, the number of finance bro undergrads

that come into my office and are like

the trouble with the GR was that some bad hombres defrauded some poor people with subprime loans

but for the bad hombres, there would have been no GR

and my intuition is that if we shot all the bad hombres, the housing bubble would have been 5% smaller but that the size of the GR may have been about the same

ocamlmycaml

anyways, I post these charts because it's worth thinking about pre-Sep 2008 and post-Sep 2008.

liquidity had declined, but you don't see catastrophic TED or equity volatility until Lehman/AIG

in early 2008, some people at the Fed were legit worried about inflation

gorbachev

so my natural instinct is to be like, disregard fraud, address structural weaknesses

geerussell

More like but for fraud the GR might have been the gR ... or maybe just the R.

That is, fraud raised the magnitude.

mrdannyocean

i'm less inclined to blame outright fraud and more inclined to blame ninja loans

and loans that just should not have been given

geerussell

ninja loans are outright fraud

mrdannyocean

are they? I'm not certain what the legality of them was at that time

there may have been a regulation that said 'hey check for income you guys' but it was not enforced at all

raising the issue of defacto vs de jure legality

geerussell

The absence of regulation in which fraud flourishes is itself a structural weakness.

ocamlmycaml

fraud is pretty broad. lot of firms didn't have a great grasp of the risk on their balance sheets. is that fraud?

roboczar

yeah i think the only way to answer this is to see if we have studies on the depth of fraud pre crisis, and figure out how to define fraud

is just being inattentive toward the makeup of your assets fraud? not sure

gorbachev

geerussell, claiming the GR would have been markedly smaller in magnitude but for fraud

is actually quite a large claim, and one that I think requires substantial evidence to support

because fraud only is relevant at the stage of "pumping up the size of the housing bubble"

geerussell

Sure, and the bigger the bubble, the bigger the bust.

gorbachev

but it's a) unclear how large the contribution of fraud to the housing bubble actually was

and b) unclear that the relationship you just suggested

namely, a linear or other such relationship between bubble size and financial crisis size - it's unclear that that is correct

For example, it's perfectly possible that you merely need a housing bubble above some threshold size to kick off a financial crisis

I mean, consider the chain of events sketched out earlier in the conversation:

Bad home mortgage lending decisions fraud ----> housing market collapse ---> a bunch of financial institutions lose their shirts on misc derivatives -----> some institutions lose their shirts ----> other banks look at their assets, begin worrying about which of their counter parties actually can pay up ----> big freeze up in trading and lending between financial firms ----> commercial paper & repo markets freeze up too slash experience what is essentially a bank run -----> across the board credit crunch at the same time industry specific shocks, like to construction at the same time, big wealth shock ----> unemployment

It's possible that a housing bubble that was -5% in terms of the amount of distortion could map into a financial crisis of the same size

since the housing bubble bursting is in many ways just the trigger for this longer string of events

geerussell

Chain: Origination -> securitization ->ratings -> sale of risky securities to risk-free buyers

ocamlmycaml

perhaps counterfactually, was there anything the Fed / Treasury could have done in mid-2008 to stem the severity of the recession?

roboczar

well, i know how post-keynesians would have answered that :stuck_out_tongue:

geerussell

They could have unclogged the housing market with mortgage cramdowns and a firm hand in foreclosures.

Instead of the combination of housing stock rot and financial damage for both borrowers and lenders as a kind of limbo dragged on for several years.

roboczar

i don't think either of those are in the Treasury or Fed's purview though?

i know it's not in the Fed's

geerussell

gorbachev, what I'm suggesting is the kind of 20%-ish year over year house price increases we had when the bubble was cresting would've been a lot smaller without fraud to feed it.

At each step of that chain.

gorbachev

This is sort of a sidebar, but I would imagine that the effect of fraud would be less on average price but more on who has a house.

geerussell

You could also see it in dramatic escalation of price-to-income

gorbachev

Like, fraud I would think largely would contribute to increases in lending at the low end of the market, while I assume average house price increases are mainly driven by gains at the top?

That sidebar being put away, my point is just that it seems perfectly possible to me that any housing bubble where prices are, say, >=X% higher than whatever than the "true" price

might be sufficiently large to trigger the chain of events that converts the housing bubble into a credit crunch

I would add I think it's super difficult to justify a dose-response relationship between housing bubble size and bust size, so to speak

I mean, imagine a counterfactual world where there are no derivatives and 100% transparency about every firm's asset ledgers

in that world, the dose response relationship makes sense - more mortgages going bust = bigger wealth effect on consumers and more assets going bad on bank ledgers

geerussell

In that counterfactual world it's more self-limiting because risk-averse asset holders aren't scooping up high-risk assets.

roboczar

NBFIs existing makes all this a lot harder and more complicated.

gorbachev

but the mechanisms that seem to have escalated the fin crisis hinge critically on how the housing market collapse

interacted with misc derivatives and with the sheer amount of uh, non-transparency of financial institution asset sheets

and completely freezing credit

mrdannyocean

especially how it ended up impacting the financial paper markets

gorbachev

Like, a housing market collapse is one thing, but if I just could look at your books and be like

oh okay, you're down 10%, I understand what your position is and what mine is

it wouldn't be so bad relative to all of your financial institutions suddenly having no clue who is a low risk counterparty or even what their own situation is

and that's sort of the crux of the credit crunch issue

geerussell

That's why a pipeline of financial alchemy where garbage goes in one end and comes out AAA from the other was such a doomsday engine.

An endless supply of risky borrowers feeding a boundless appetite for risk-free assets.

gorbachev

you need the housing bubble to trigger it, but it seems plausible that several different sized housing bubbles could have mapped into a similarly sized financial crisis

because on some level the size of the fin crisis was less about # subprime loans or whatever and more about the interlinkages between financial firms, the poorly understood structure of derivatives, etc

and i mean, you don't need fraud for that to blow up basically. like people were dumb enough to believe in the garbage ----> AAA alchemy, even without fraud

ocamlmycaml

people acted as if MMFs weren't runnable at all

gorbachev

As a firm, I could honestly have sold you trash and if the contract around it was fancy enough, you were all in

Don't get me wrong, there are good independent reasons to want to crack down on fraud, but it doesn't seem to me like a very important part of the GR story. not enough to send us from GR ---> R I mean

geerussell

It's kind of important as gasoline on the fire.

gorbachev

It's not clear to me that that analogy is appropriate.

geerussell

At the start of the pipeline a steady and growing supply of new origination was needed. The analogy applies to the extent that fraudulent origination kept it fed with new loans and rising prices.

A lot of corner cutting was needed to move all that volume through into securitization. That corner cutting also fed the size and made it harder to clean up afterwards.

gorbachev

I mean, but what's the evidence on a) how many loans were issued that would not have been, but for fraud and b) on whether or not that difference in quantity of loans made a difference in terms of translating the housing bubble into a financial crisis

geerussell

Fitch gives a nice 2007 canary in the coal mine overview of how fraud factored in here

http://big.assets.huffingtonpost.com/FraudReport8Nov07Fitch.pdf

ocamlmycaml

I think this where a model would be helpful. An "overdetermined system" if you will

geerussell

The FCIC report also gets into a number of relevant issues, including appraisal fraud

https://www.gpo.gov/fdsys/pkg/GPO-FCIC/pdf/GPO-FCIC.pdf

gorbachev

okay, so, can we just step back a minute on that fitch report and think about what they're saying

3 million loans, 16k apparently w/ fraud. within an ultra small sample of subprimes, they claim a high fraud rate

but then to get to their really big headline #s on subprime residential mortgage backed securities

the "fraud is 25% of the underperformance" thing, they're basically attributing the residual in their regressions to fraud

So if we sum that up, we have a ratings agency trying to argue that it's not that their subprime MBS ratings were wrong - it's just that they were right but that they got duped by fraudsters

I'm not saying my faith in the financial industry is low, but given that it seems rating agency conflicts of interest, among other issues, contributed to the brutally bad ratings of misc derivatives

geerussell

Ratings agencies were complicit but at a different stage far removed from origination.

gorbachev

I'm not totally sure I buy a rating agency trying to pin their failure on fraud by random poor people putting the wrong shit on applications while bankers turn a blind eye.

Like this just seems like the most self-interested, pass-the-buck style answer

and given the evidence is in large part them interpreting a residual.........

geerussell

blind eye/actively pursuing it. If no regulator is going to rap you on the knuckles for it, it's very profitable.

gorbachev

also wasn't the problem not that like, subprimes were crappier than expected on the margin

and more like, people literally thought subprimes were 0 risk if bundled together?

geerussell

That was a problem too and where ratings agencies were complicit.

gorbachev

like, maybe fraud meant that each asset's return should be downrated by x%, but fraud isn't responsible for why those risks were correlated

and the correlation across risk is what made CDOs and stuff risky as hell, and missing that correlation issue was big......

geerussell

They kind of go hand in hand. Fraudulent appraisals support garbage loans that feed risky MBS where the risk is obfuscated.

I don't buy the "random poor people" story either, it took actively dishonest lenders to keep it going. From the FCIC report for example:

Some real estate appraisers had also been expressing concerns for years. From 2000 to 2007, a coalition of appraisal organizations circulated and ultimately delivered to Washington officials a public petition; signed by 11,000 appraisers and including the name and address of each, it charged that lenders were pressuring appraisers to place artificially high prices on properties. According to the petition, lenders were “blacklisting honest appraisers” and instead assigning business only to appraisers who would hit the desired price targets.