It’s rare that I find fault with the the Financial Times, and even more uncommon with Gillian Tett and Paul Davies, who are two of their most seasoned and insightful journalists. Nevertheless, they have bitten off more than they can chew in “Out of the shadows: How banking’s secret system broke down.” The piece isn’t bad, but it fails to deliver on its promise. The article fails to define what the “shadow banking system” is, much the less explain why it is under stress now. And iit also mischaracterizes how we got where we are.
Before we turn to this article, we need to focus on nomenclature. One factor that may have gotten in the authors’ way is that different people may use the same expression to signify different things, which leads to imprecision and sometimes error.
Shortly after the credit crisis hit in August, there was considerable discussion at the Federal Reserve’s Jackson Hole conference of why this crisis was different. In particular, James Hamilton’s presentation did a very good job of summing up where the problem seemed to lie (and it’s troubling that his observations still aren’t getting the traction they deserve):
What has happened over the last decade is that a variety of new institutions have evolved that play a similar role to that of traditional banks, but that are outside the existing regulatory structure. Rather than acquire funds from depositors, these new financial intermediaries may get their funds by issuing commercial paper. And instead of lending directly, these institutions may be buying assets such as mortgage-backed securities, which pay the holder a certain subset of the receipts on a larger collection of mortgages that are held by the issuer. Although the names and the players have changed, it is still the same old business of financial intermediation, namely, borrowing short and lending long.
There are a variety of new players involved. The principals could be hedge funds or foreign or domestic investment banks. Others could be conduits or structured investment vehicles, artificial entities created by banks, perhaps on behalf of clients. The conduit issues commercial paper and uses the proceeds to purchase other securities. The conduit generates some profits for the bank but is technically not owned by the bank itself and therefore is off of the bank’s regular balance sheet.
This system has seen an explosion in recent years, with the Wall Street Journal reporting that conduits have issued nearly $1.5 trillion in commercial paper. Their thirst for investment assets may have been a big factor driving the recklessness in mortgage lending standards, as a result of which much of the assets backing that commercial paper have experienced significant losses.
Without an adequate cushion of net equity for these new financial intermediaries, and with tremendous uncertainty about the quality of the assets they are holding, the result is that those who formerly bought the commercial paper are now very reluctant to renew those loans, a phenomenon that PIMCO’s Paul McCulley described at the Fed Jackson Hole conference as a “run on the shadow banking system.” I heard others at the conference claim that there might be as much as $1.3 trillion in commercial paper that will be up for renewal in the next few weeks, with great nervousness about what this will entail.
In some cases, these intermediaries have lines of credit with conventional investment banks on which they will be drawing heavily, which will cause these off-balance-sheet entities to quickly become on-balance-sheet problems. Their losses may severely erode the net equity of the institution extending the line of credit. How big a mess will this be? I don’t think anybody really knows for sure.
This discussion is succinct and clear. It defines the nature of the problem (borrowing short and lending long) and gives a clear definition of what markets and players are involved (commercial paper and vehicles that use it for funding). And indeed, what has troubled regulators is the disruption resulting from the implosion of the asset backed commercial paper market. Banks are under stress because they are the other big source of short-term funding. They are hit from multiple sides: companies and entities drawing on lines of credit that no one anticipated would be used on a large scale, all at once; providing funding to affiliated entities that they thought they could cut free if they needed to but now realize they can’t; finding that funding in any form, short-term, medium term, equity, is costly and scarce.
Contrast Hamilton’s overview with the FT’s article, which never defines what the “shadow banking system” and as we’ll discuss, even lumps in a major type of instrument that doesn’t belong on the list. The story is surprisingly flabby and unfocused, although it does marshal some useful data:
Yet while investors are scrutinising some of the industry’s best-known names, a spectre will be silently haunting events: the state of the little-known, so-called “shadow” banking system.
A plethora of opaque institutions and vehicles have sprung up in American and European markets this decade, and they have come to play an important role in providing credit across the financial system. Until the summer, structured investment vehicles (SIVs) and collateralised debt obligations (CDOs) attracted little attention outside specialist financial circles. Though often affiliated to major banks, they were not always fully recognised on balance sheets. These institutions, moreover, have never been part of the “official” banking system: they are unable, for example, to participate in Monday’s Fed auction.
But as the credit crisis enters its fifth month, it has become clear that one of the key causes of the turmoil is that parts of this hidden world are imploding. This in turn is creating huge instability for “real” banks – not least because regulators and bankers alike have been badly wrong-footed by the degree to which the two are entwined.
The article is hamstrung by failing to establish what this parallel banking universe is, beyond citing SIVs and collateralized debt obligations as examples and later “vehicularized finance.”
But in fact, what the FT piece does is conflate the sort of problem that Hamilton defined crisply – that of new non-bank entities acting like banks by borrowing short, in this case via commercial paper, and lending long – with another financial development, securitization. The stresses arising from CDOs are not a shadow banking system issue as defined by Hamilton, except in those cases that CDOs were structured to use commercial paper for a “super senior” tranche. As we have discussed, this was unnecessary, dopey, greedy, and could be expected to blow up, as it has. But so far, the only issuer known to have used this, ahem, innovation is Citigroup, although they did it on a grand scale.
So CDOs are part of the “shadow banking” problem only to the extent that they issued commercial paper; by every indication, this took place only to a limited degree in a very big market. Yet the FT piece would lead you to believe that CDOs are every bit as much a culprit as SIVs.
That isn’t to say that CDOs aren’t creating stress in the markets. But the “crisis in the shadow banking system” talk is all about a run on non-banks. The scrambling for a resolution of SIVs’ inability to roll maturing commercial paper is a classic manifestation.
But the CDO trouble (which by the way is primarily hitting the very large debt trading houses, not commercial banks generally) operates by the same mechanism that subprime, credit card, home equity loans and commercial real estate are, via credit losses. The financial firms are taking writedowns that are large enough that they are reducing equity. This has two effects: it reduces balance sheet capacity, and it also tends to produce new-found caution. Thus the intermediaries typically become more stringent in their credit policies, beyond what is warranted by their equity losses. In fairness, the excessive conservatism may be warranted, since it isn’t clear yet that all the dead bodies have been found (i.e., there may be further balance sheet shrinkage).
But the FT is clearly confused about where the problem lies:
By any standards, the activities of this shadow realm have become startling. Traditionally, the main source of credit in the financial world was the official banks, which typically forged business by making loans to companies or consumers. They retained this credit risk on their books, meaning that they were on the hook if loans turned sour.
However, in the past decade, this financial model has changed radically. On the one hand, banks have increasingly started to sell their credit risk to other investment groups, either via direct loan sales or by repackaging loans into bonds; at the same time, regulatory reforms have permitted the banks to reduce the amount of capital that they need to hold against the danger that borrowers default.
Huh? Where have the writers been? Banks have lost market share in the credit game to investment banks since 1980. If you are going to claim that the culprit is disintermediation, it isn’t a ten-year old phenomenon. And recall that in the last credit crisis, the LTCM workout, the Fed called on the Wall Street firms that were its creditors, evidence that the world had already changed by then.
Now there is likely a case to be made that the increase in systemic leverage makes a difference in degree into a difference in kind, that with the uses of leverage-on-leverage (hedge funds which are geared investing in CDOs which have a lot of embedded leverage) that a lot of players have been put in the position of facing large price swings (more accurately, declines) that they assumed they’d never see, and then facing unexpected demands for liquidity (think of the redemptions at quant hedge funds, which ironically were not players in the sort of structures that this article discusses; they stuck with instruments that were liquid). But hedge funds have taken to freezing withdrawals if things really get dire to assure an orderly wind-down.
Consider this comment:
Satyajit Das, an author and derivatives industry expert, cites an example where just $10m of real, unlevered hedge fund money supports an $850m mortgage-backed deal. This means $1 of real money is being used to create $85 of mortgage lending – credit creation far beyond the wildest dreams of high-street bankers.
The writers fail to frame why equity is necessary, and that omission again clouds the discussion. Banks need liquidity and hence need a buffer primarily due to their funding mis-match. Their depositors “lend” short, and so banks need a cushion both to deal with cash needs and possible losses due misjudging or mismanaging credit or term risks (presumably, they priced the credit with certain assumptions in mind). In theory, if you had a perfect asset-liability match and had priced the credit correctly for credit losses, you’d need no equity. The alleged better ability of investors to tailor their assets (the debt instruments they buy) to their liaiblity requirements is one of the drivers of bank disintermediation (the other is that investors don’t face the cost of deposit insurance, a non-trivial item).
That’s a long-winded way of saying that disintermediation and the resulting increase in overall leverage could readily be depicted as an accelerant of the shadow banking crisis, but the authors don’t make that case.
Despite this critique, I nevertheless encourage you to read the article; it has useful comments and factoids, despite the flaws in the overall argument.
I read the Tett and Davis piece and thought it stank. Thanks for the lead on Hamilton. I agree the CPA profession let the public down. I am a CPA and am disappointed to say the least with my fellows. The first treatise I ever saw about the dangers of borrowing short and lending long appeared in Holland in 1587. There is nothing new here.
I think you are being a bit severe – its a well written article, that admittedly glosses over detail several layers down, but is not fundamentally misleading. It’s a difficult and complicated subject to summarize.
One comment on what you’ve written:
“In theory, if you had a perfect asset-liability match and had priced the credit correctly for credit losses, you’d need no equity.”
Banks price credit by building a factor for expected credit losses into the spread, and then providing capital for unexpected credit losses. In other words, the credit risk that must be supported by capital is the potential deviation from the expectation. Of course, if banks priced ‘correctly’, there would be no risk, because actual credit losses would always equal expected credit losses. Credit losses would be a cost but not a risk. This is impossible in a risky world.
I think this notion of leveraging up 80 times is flung around quite loosely, as is the notion of ‘real money’. I think this is disorting the analysis of true macroeconomic exposure.
The effective leverage factor may be 80 times, but that also means the underlying exposure applicable to that security is 80 times as small as what it would be with no leverage. The probability of loss of course is much higher, but the exposure in terms of maximum loss as determined in the nominal size of the exposure is much smaller. You may be leveraged 80 times, but you can’t lose more than 1/80 of the total nominal underlying exposure.
It’s not just a matter of maturity transformation, but of the creation of credit “liquidity” in excess of investment opportunities in the real productive economy; which is to say, of course, the leveraged build-up of asset inflation in the financial sector out of any balance with any possibility of realizing profits and incomes in the real economy. And underlying that escalating leveraged financial asset inflation is the stagnation of wages and salaries and the over-capacity/off-shoring of the real production economy, which is being offset by the growing financialization/financial disintermediation of the real economy. In other words, leveraged asset inflation is at once caused by and the cause of mal-distribution and excess inequality (functionally speaking) of income and thus shortfalls in effective demand. Reflating the deflation in the real productive economy through loose monetary policy has only exacerbated the problem, which speaks to the limits of a laissez faire overreliance on monetary policy, when fiscal (and regulatory and public-administrative) policies are required to counter the underlying imbalance. (And, just to inject an explicitly political note, it seems to me that Rupublican policies/ideology result in a recurrent mismatching of fiscal and monetary policies, with the latter forced to compensate for the utter deficiency of the former).
I don’t understand anonymous 8:12 AM’s comment about the “macroeconomic” effect of leverage. Sure, the leveraged structure might be built-up over assets amounting to just 1/80 of the value in the real economy, but the excess financial profits accruing to leveraged structures must be drawn from somewhere, (though I suspect that that somewhere often amounts to future financial losses in the financial system offsetting already booked “profits”). And when the leverager loses the 1/80 of his actual equity, other parties are left to take further financial losses in the unwinding of the leveraged structure and the deflation of the value of its assets. The “macroeconomic’ effect of excessive leverage is massive financial asset deflation and debt overhang in its eventual unwinding, which sooner or later must disrupt the functioning of the real economy.
I agree with Halasz. Some of the profit from businesses like leveraged buyouts comes from bankruptcy law abuse. The LBO guy buys 10 companies. He levers them up. Two go bust. Who bears those losses? It’s heads the LBO guy wins, tails the creditors lose. The finance industry is much the same way in creating “redistributional” gains and losses.
john c. halasz and independent accountant:
I don’t disagree.
On the contrary, independent accountant’s point on redistributional effects is what I’m getting at. And the LBO guy creates a long call option on underlying assets, and leaves others with the debt short put option. Total underlying losses get redistributed through this capital structure.
The proliferation of such strategies may increase overall original financing supply and stimulate the creation of bad assets in the first place, but it can’t further magnify the losses generated by them, apart from the operational and legal cost of settling the result.
Financial asset deflation and debt overhang are the result. But financial asset deflation would also have resulted had the same underlying assets been capitalized with 100 per cent equity. There is rightful outrage that so much of these assets are now capitalized with debt. But that doesn’t increase the total size of losses per se.
Credit default swaps are a form of off-balance sheet financial intermediation. The result is a 0 sum game of those betting on reference assets. Some win; some lose. Some use bets to hedge but lose on their counterparty. It’s still a 0 sum game at the margin. At the more basic level of on balance sheet intermediation, a leveraged bank that buys shares in another leveraged bank has multiplied the effective leverage on the ultimate underlying assets, but hasn’t increased the total exposure generated by those assets.
There is a toxic chain of such intermediation and leverage throughout the system now. I don’t underestimate the ugliness of chain reactions of losses in this context, but apart from settlement friction, losses of value and income on the underlying assets can’t really be magnified simply due to intermediation and leverage per se.
My point is that I think there’s a tendency to associate things such as SIVs investing in CDOS that invest in subprime, or such as $ 40 trillion in CDS (or whatever the number is), with net incremental financial and real losses to the economy as a result of these leveraged structures per se, which is not really correct. The ultimate losses are generated and mostly defined by the failure of the underlying subprime mortgages in this case.
I agree with halasz but am puzzled by anonymous. Anonymous, if the system worked as you said, the value of defaulting/losses on subprime mortagages would= losses that have resulted so far, which is blatantly not the case. Just as there’s the “multiplier” effect in the larger macroeconomic picture, leverage provides the same effect in financial intermediation. Though theoretical physics says the amount of energy is the same, just dispersed or transformed in different ways but risk is slightly different , risk redistribution doesn’t mean that each of the downstream investors/intermediaries face the same amount of risk when aggregated, particularly when downstream investors/intermediaries are themselves interconnected in a spider’s web sort of way. This means higher credit risk for each downstream intermediary, part of which is created by the very act of transfer of x% of risk from the party A to downstream party B . It is risk reduction for A but risk increase for B and not necessarily in the same amount . What would further complicate matters is if the amount of risk B faces is predicated on the performance or risk profile of A. Imagine what happens when C’s involved.
Not sure if that made matters clearer or worse. Ok, will be the first to admit my dumbness, do provide a counter-explanation if that was way whacky
On a more general note, Yves is I believe speaking to the neoliberal project’s decades long financial liberalization (deregulation) from which, particularly post-1990, grew an expansive global system of non-bank banks. Very evidently the competitive and otherwise interactings of these and the traditional banking sector produced a self-financing* and progressively more uncontrollable hybrid or, as mentioned in 2001, ‘nuclear credit fission’.
There remains a general failure to recognize what has developed, with most still under the spell of the traditional, a ‘spell’ from which eminates demands that governments and central banks do what they are no longer able to do, effectively mitigate.
*’Self-financing’ should not be taken to mean autonomous but instead the ability to transform liabilities into assets types in more through and through fashions giving the system as a whole a certain ‘evergreen’ quality while simultaneously creating greater tension between itself and its last instance basis, production capital (which, owing to differential rates of return, has also been increasingly financialized). Or, J. C. Halasz is perfectly correct to counterpose the ballooning of claims and the production economy, which have come to form a terribly imbalanced unity.
I don’t understand your point about losses to date equalling total losses. I’m not saying that.
There is no ‘multiplier’ effect. Leverage doesn’t do that. Suppose a house is funded 100 % with equity in one case, and is leveraged 20 to 1 in another (95 % debt; 5 % equity).
Total losses to total funding in either case can’t exceed the original cost of the house. Total losses to total funding are the same for a given loss on the house value, no matter what the funding (apart from legal and other frictional costs associated with distributing the losses).
Leverage makes no difference to total losses absorbed by the financial system.
If one of the lenders hedged his position with a credit default swap, the loss is transferred from the lender to the counterparty. No net change to losses absorbed by the financial system. If the counterparty fails to deliver, the loss is transferred back to the lender. No net change to losses absorbed by the financial system.
And so on…every step of the way can be analysed like this … there is no increase in total losses to the financial system due to either the compounding of financial intermediation or the complexity of capital structure (including leverage). Interconnections don’t create additional net risk for the system. They only transfer it.
All financial asset risk is a transfer between two counterparties. This is the case whether it is cash or derivative risk, because cash risk is the sum of risk free funding and an embedded derivative position. It is generally accepted that derivative risk is a 0 sum game. E.g. a call option transfers risk from the buyer (long) to the writer (short). So cash risk is also a 0 sum game. E.g. cash credit is the sum of risk free funding and a short put option position. A ‘leveraged’ risk position is simply a combination of equity (embedded long call option on the underlying asset) and debt (embedded short put option).
If all financial asset risk is a 0 sum game, then the only net risk left in the system is the underlying real asset risk (where real asset risk sometimes incorporates human capital risk – i.e. the ability of the owner to generate employment income). That net risk is transferred from the beginning of the chain (the value of the house) to the end of the chain (the final financial asset holder who has only his own equity position to absorb his financial asset risk, rather than pass it on through yet another financial liability).
I think that you are confusing particular or idiosyncratic risk with systemic risk, as though the latter is merely an aggregation of the former rather than an essentially unquantifiable, dynamic and unintended synergism. Or simply a whole greater than its parts.
anonymous @ 11:57 p.m.
I’m familiar with portfolio theory, in which total risk is less than the sum of all risks. I’m familiar with systemic risk, and there is no corresponding theory I’m aware of that suggests the reverse. I’d be interested if somebody can provide a simple conceptual example. (Analogies don’t count as examples.)
rather than portfolio theory try actuarial VaR and superadditivity for some examples — the notion that total or systemic risk would be same or less than the aggregate of particular risks is one of the reasons we are where we are today as well as the dovetailiing with a commonplace neoclassical fixation on the micro.
anonymous 10:29 PM:
Yes,in principle, all economic exchanges are of equal values, and all attached risks sum to zero. The question is then why such exchanges and transfers of risk should occur at all. Your point can be inverted by noting that, given a pool of risks underlying a structured financial vehicle, the total risk, whether accurately modeled or anticipated or not, remains the same, regardless of how it is sliced and diced or redistributed. There is no net gain from the “enhancement” of credit; the only thing that really counts is how accurately the base-line risk of the original extention of credit is gaged in terms of the outcomes of underlying production in the real economy. The “fundamentalist” point here is that all returns to and of financial assets ultimately derive from productive surpluses in the real economy, from long-run and illiquid investment of real productive capital and the revenues realized from the sale of output. (The corollary of maturity transformation is the forward projection of economic outcomes.) And there is some indeterminate limit as to the debt-load that can be really carried and realized by the real productive economy. Which is to say that either the hyper-extention of the financial economy is redundant and superfluous,- (in which case why the tremendous investment of labor and capital resources in it?),- or it is a misguided estimation of future returns from the outcomes of the real productive economy,- (since fees are extracted with each financial pass-through from the initial assessment of risk undewriting the initial extention of credit, thus underpricing risk and distorting economic allocations),- or it is a rent-extracting activity, drawing off productive surpluses from the real economy and minting profits without corresponding costs of production. In any case, but especially the last, there are costs drawn from, but then displaced back onto the real productive economy, which may well be more than just “frictional”, distorting and disrupting the balanced reproduction, development and growth of the real economy, which through interlocking negative feedback loops, magnifies losses to the latter beyond the immediate instances of financial defaults. Money of any sort, after all, is just a semiotic instance and not a real “thing”, merely representing- (imperfectly, due to the embedding of both costs and information in production systems)- the real “things” that are produced and exchanged in the underlying real economy. And it is the provisioning, deepening and improvement of those real “things” and their means of production that is the point of any modern economy. If the self-referential operations of the financial economy are drawing of surpluses excessively from the real economy and distorting the distribution of both claims and risks in the real ecomony required for its balanced reproduction, (not least of which is some optimal balance in the distribution of income, qua claims on the real distributable surplus product, between labor and capital, wages and profits, required to sustain adequate effective demand), then it is, indeed, producing and displacing real losses greater than the initial losses on defaulted underlying assets.
john c. halasz
Thanks very much for your response, which I’ve only just seen due to my unfortunate scheduling.
You’ve written something quite extraordinary. I’m saving it.
I know now I believe at least 95 per cent of what you’ve written and almost certainly approaching 100 per cent with a little more time.
I don’t think our thoughts are inconsistent. I attempted to segregate the result of pure financial intermediation and leverage effects – not the cost or effectiveness of human labour effort expended to create a (new) financial profile against underlying real assets.
The cost of financial ‘creation’ in this sense may not justify the result in terms of financial disposition of the realized risk.
I apologize if I’ve underestimated your message but I must flee.
Thanks again. Great thoughts and words on your part.
anonymousDecember 17, 2007 10:29 PM
>>>I don’t understand your point about losses to date equalling total losses. I’m not saying that.
erm, not saying that either, meant that as the losses on each of the elements on the chain aggregate, they do not net off or get to the total you might have expected in the beginning. I think halasz has come up with the theoretical explanation.
Interconnections might impact negatively on risk profiles partly due to credit risk but partly due to the interdependence, for example when asset y leaves A for B but the risk profile of asset y continues to be affected by the risk profile of A or the performance of quantities of asset y still remaining at A or sold by A to other parties like C. Not sure if that made sense, but simply put, as each element of the risk/product chain passes the product and some part of the risk on, it’s creating greater uncertainty and that amplifies the risk inherent in the chain, by the time it gets to the investor, the risk factors have multiplied and is almost too convoluted to be comprehended.
When risk is transferred, it affects perceptions of risk and therefore, may affect credit risk more than what is mathematically warranted, take for instance credit ratings and default swaps (which tecnically reflect perceptions of the possibility of default ) which are arguably more vulnerable to being affected by perceptions, the mistrust of ratings issued by credit agencies is arguably what is causing investors to turn risk adverse. That is what I term as the multiplier effect, for lack of technical knowledge on my part.
I know what you’re saying it’s just a difference in opinion I guess.
Actually, part of this response was already on my blog , looking forward to your response either here or on foesskewered.livejournal.com
I’d have to think through a specific example to see how such a risk factor might be clearly separated out. I know it seems intuitive, but there should be a simple way of demonstrating such an effect and that’s not obvious to me.
There is such a temporary effect in marking to market or model. Actual financial asset losses aren’t known until the mortgage defaults and the residual asset value is determined (unless the financial asset is sold to somebody else, who then has that problem). Until then, banks are making estimates of the value of their own positions based on an estimate of the value of the underlying realizable real asset value. This is the case for any loan loss provisioning. Current estimates may turn out to be too low requiring further write-downs on financial assets or too high allowing loss recoveries.
Not too sure about examples but could there be one in the chain that goes from, for example, the mortgage (originating in say, a mortgage specialist), going along to maybe a commercial bank/bond issuer, who slices that and incorporates it in ABS or CDOs ,or issues commercial paper with the “risk profile” as assessed by credit agencies , bought by funds who are in turn invested in by pensions or other investors. Of course, in the process there are the bond insurers and possibly brokers.
The perception of risk then goes from mortgage specialist all along those paths and because of additional risk factors (not just default risk, but for example, the risk that credit ratings are no longer realistic etc), ultimately even short term funding instruments that have no connection with the mortgage industry are seen as “tainted” partly ‘cos uncertainty has now infected each part of that chain.
does that make sense to you? it’s a very abbreviated explanation but, sorry, gotta catch some lunch before the time’s up!hope the example makes some modicum of sense!
That does make sense. I think of it as many different financial assets whose values derive from the original mortgage. As uncertainty is transmitted throughout the chain, the loss in financial asset values may ‘overshoot’. Financial assets will trade and sellers will lose money while buyers put those assets on their books at lower prices. But the ultimate value of those assets can’t be known until the mortgage actually defaults and the residual value of that asset is determined. Then the residual value of all derived assets can be determined as final relative to the performance of the original mortgage. Some sellers would have lost more money than ‘necessary’ relative to this final value, offset by some buyers having purchased assets at a discount relative to this value. There is a 0 sum exchange of ‘error’ gains and losses according to the earlier estimates of value that corresponded to trade prices.
Apologies, anonymous, not sure if you still track this particular post, let’s just say there was a gingerbreadman incident, which I’m still recovering from
Iunderstand what you’re stating, that’s what the the overall macro-economic equations (uncannily like accounting identities/equations) dictate but it’ll be hard to fit the insurers’ eg MBIA into the equation, sometimes, imbalances are rarely addressed/netted off till the long term and as an economics tutor of mine used to say, in the long term,we’re all dead