Pers Sjoberg and Susan Hinko of TriOptima argued in Financial Times this week that OTC derivative markets aren’t liquid enough to be subject to real-time transparency requirements. But what does liquidity have to do with the social benefits of price transparency?
The correct comparison is not futures or option markets, but other illiquid markets like the corporate bond market. And it’s far from clear that real-time transparency has been detrimental to the corporate bond market — in fact, the most recent news has been that the corporate bond market is thriving.
The authors argue:
[In] highly-liquid, standardised markets, real-time reporting and price transparency are meaningful concepts. When these concepts are applied to the OTC markets, adjustments should be made, especially given the fact that many of these low-frequency transactions have large notionals and need to be hedged discreetly to minimise hedging costs.
The whole goal of price transparency is that it should be impossible for anybody to “hedge discreetly.” That’s the point. Every transaction shows up in the market price. If I do a bond transaction sequentially for two different accounts, I’m likely to see a price movement from one transaction to the next — and I trade in very small quantities. But this is simply the inherent nature of trading on a functional market where my earlier purchase (which was completely unpredictable before I sent it in) shows up as having a price effect after it has been made public. Price movements, like this, are the very purpose of transparent markets.
The authors continue:
However it is important for the new rules to reflect the realities of the market so regulators and the public have access to meaningful information. Intra-day reporting will place systems and cost burdens on institutions already restructuring to meet the demands of new legislation and regulation, and will not yield significant results to the public.
I don’t understand why purchases like my first purchase of an illiquid bond are classified by the authors as not “meaningful information.” That’s a decision that the dealers on the market make when they post their bids and offers. If a transaction is not meaningful to them, it will not shift the price, and, if it is meaningful to them, it will.
We can all understand that it is more beneficial to financial institutions for them to be able to trade with more information than the rest of us, but it is far from clear that protecting their intra-day information is in the public’s interests. If the only costs are structural, because financial institutions will have to set up real-time reporting systems, it seems to me to be in the public’s interest to create a standard that says that such systems should be set up for just about everything that financial institutions trade. Thus, when regulators want real-time reporting of a new product, it should take the financial institutions days, not weeks, months or years to provide.
We live in a society where every pack of gum is bar-coded and consumers can do cross-store price checks on their phones. Why has the financial industry chosen to maintain back office procedures that are practically medieval — even as they trade in multi-million dollar products?
Synthetic Assets 
But what does liquidity have to do with the social benefits of price transparency? […] The whole goal of price transparency is that it should be impossible for anybody to “hedge discreetly.” That’s the point. Every transaction shows up in the market price.
Sure. But in the situation in question, is it possible that the market price will have dubious ‘social benefits’?
One wishes to hedge discreetly because the opposite of that is this: everyone knows you have a bunch of risk to hedge, so they get out in front of you and run the price up. What are the ‘social benefits’ of front-running?
You don’t go into a car dealership and announce “I really need and want this car ASAP”. Should you be forced to do so (i.e. to publicly announce to all car dealers that your old car was totaled and you need a car for work on Monday) in the interest of price transparency’s social benefits? The resulting price you pay for your ‘hedge’ would presumably be distorted.
If you agree that prices can be distorted by front-running and that this can have negative social benefits (or do you?), then liquidity – namely lack thereof – could very well have everything to do with the social benefits of price transparency. Am I wrong?
Best,
First of all it’s not really clear to me how your “front-running” works. Let’s say x is the normal trade size on this market and there are 5 dealers who are obliged to honor their posted bids/offers for a quantity x. Thus if I place an order for 5x, it doesn’t move markets (as far as I’m concerned), I just get filled. The problem arises when I want to place an order for 50x. If I place it as a market order, I’m guaranteed to get poor prices — but I don’t think that would be due to front-running. More likely I’ll post it as a below market limit order, get the 5x at market price and, if I set the price right, can get the other 45x filled in a reasonable period of time. Of course, the other market players can choose not to fill me to see whether I’m going to change my limit price, but I always have the right to hold my position and demand the price I want.
As I understand it, what you call front-running can only occur when you’ve made a strategic error that puts you in a position where you so need the asset that you can’t wait for a fill at a below market price, but have to drop your limit price even lower. (Of course, it’s also possible to try trading in a sequence of orders, but that already presumes that you have time and don’t really need the hedge quickly — and it’s far from clear that sequential trading will get you better prices on a transparent market.)
Secondly, I think of the issue as wanting to make sure that the externality imposed on the market by the decision to, for example, buy as a hedge is internalized by the buyer. If the buyer has mismanaged their positions so badly that they have a “need” to make a large sudden purchase that will move markets, then the costs of the subsequent price movement should be born by the buyer. It is precisely because of the immediacy of the buyer’s “need” to draw excessive liquidity from the market that there’s an issue of an “excessive” price change — but if the need is “immediate” there’s really no way to claim that the price movement is distorted.
If I’m with someone in a car accident in country without ambulances and I need transportation to the hospital ASAP, the fact that the passing car owner can extort a vast amount of money from me is hardly a sign that markets aren’t working correctly. The market price is a function of my need — and if I’m Bill Gates and the person with the car is deeply impoverished maybe $1 million is not just a market price, but even a fair price, for a ride that in other circumstances would cost $5. So I’m not even sure I understand in what sense “front-running” distorts the market.
In transparent, but illiquid markets, it is crucial for each firm to manage their positions so that they only very rarely “need” to hedge on a large scale.
I don’t understand what social benefits you see in encouraging firms to manage their books so that they regularly face a “need” to hedge discretely. That strikes me as an inherently destabilizing situation, because every few years a group of big firms will need to hedge the same way at the same time and will cause a crisis.
Surely it’s better to make the firms face the costs of hedging in the transparent market on a daily basis, precisely to discourage them from taking positions that are likely to require sudden market moving hedges.
Thanks for the response.
In some sense we are just talking apples and oranges, or different languages. ‘Market orders’/’limit orders’? These exist in the equity world, but are not even in use as such in certain markets, i.e., many of the ones under discussion. Similarly in the large-hedge scenario you rejected out of hand the concept of doing it in multiple pieces, when to me that was part and parcel of what ‘hedge discreetly’ meant – when I read that phrase, it’s precisely what came to mind. Overall, it appears you’re tacitly/unconsciously? applying principles/conventions you know about that hold in liquid markets in trying to rebut the above claim about illiquid ones, in which (the whole point was) such principles don’t necessarily apply. Essentially, this means you’re changing the subject and failing to address the actual claims that were on the table.
On the problem of needing to do a large hedge: it seems you’re missing the fact that new-risk does get originated (or, transferred). I think I would see what you’re trying to say if there were a fixed, finite blob of risk in the world just getting sloshed around gradually from here to there. Then, the only hedges anyone should ever have to put on would be to existing, slowly-changing positions, and in some ideal Platonic sense everyone would always do their hedging incrementally, in gradual, delta-infinitesemal bits. Yes, if that were a description of reality, then no one would ever have a valid reason to ‘need’ to put on a large hedge, and it would always be ipso facto evidence of mismanagement. But of course, that’s not reality. Specifically, what if someone just did a new $X00 million deal? What about new deals, and end buyers/clients? New deal = new risk origination/transfer = new, unanticipatable (at least, not well in advance) need to do large hedge. I don’t see how you can blame or fault anyone for this or characterize it as irresponsible. Unless no one should ever do new deals? Similarly, there can be ‘events’ (court rulings, regulatory overhauls…) leading to the need for a large hedge, which you also ignore. If I accepted all your premises I’d really have to start to wonder why any financial transactions ever took place at all – it seems like in your ideal world, there wouldn’t be any 🙂
On the ambulance analogy: Here’s the problem. Your whole claim here is that we (i.e. lay people, the rest of us, observers) somehow need or will benefit from price transparency in these markets. But (as on a previous post about this) you haven’t articulated why you think so. Of what social benefit would it be to anyone else to have it published/broadcast that Bill Gates paid a guy $1 million for a ride to the hospital in Somalia? On one level I am wrong (or at least being verbally lazy) to refer to this as a price ‘distortion’ – in a theoretical, free-market sense there’s no such thing, as long as the transaction was voluntary; the real point is just, how will the rest of us use or benefit from the information-content of that price information, exactly? Because if we won’t, then what’s the point?
Basically, the argument that the rest of us need and will benefit from price transparency is practically self-nullifying when you attempt, as you are here, to apply it to illiquid, chunky markets that the rest of us are, by definition, not participating in.
best
‘Market orders’/’limit orders’? These exist in the equity world, but are not even in use as such in certain markets, i.e., many of the ones under discussion.
But I view the movement to market transparency as a movement to force these markets to adopt certain norms like posted bids and offers (which should function legally as offers for a quantity x, that are enforceable contracts when accepted). Thus in my view one of the purposes of transparency is to make things like market and limit orders the more sensible way of doing business. There’s no reason an “illiquid” market can’t function in these terms (see the options markets, which for any given contract is often extremely illiquid).
New deal = new risk origination/transfer = new, unanticipatable (at least, not well in advance) need to do large hedge.
Alright I can grant that there may be some costs (that are small relative to the benefits) to transparent markets, because it would be cheaper to hedge “in the dark.” But I don’t see why the new deals shouldn’t internalize the cost of the hedge. If the transparent market is really expensive, I would just expect more firms to be brought in at the deal-making stage (rather like a syndicated bond issue) — and I don’t see anything wrong with this. In any case these costs seem small relative to the benefits to me.
how will the rest of us use or benefit from the information-content of that price information?
I tried to make the benefits clear in the post above. The whole point is to put an end to over-reliance in the financial industry on some theoretic ability to delta-hedge, which in practice does not exist because all markets can only ever be temporarily liquid (see 2008). Because delta-hedging can consume vast amounts of liquidity, it is important to get the financial industry into the habit of paying for the costs of liquidity provision in illiquid markets. In short, the point is to reduce the origination of risk, for the purpose of reducing reliance on liquidity that is guaranteed to evaporate just when its most needed and thus reducing the likelihood of financial crisis.
The benefit of reducing the likelihood of crisis far outweighs any costs created by making it more expensive to originate risk. After all in 2005-2007, it’s abundantly clear that far too much risk was originated. To me the crisis was proof that it is essential to change our financial infrastructure to curtail the origination of risk.
In short, the fact that you think transparent markets will reduce risk origination seems to me to be an argument in favor of transparent markets doing what they’re supposed to do.
It’s not clear to me how ‘market transparency’ would ever make a limit order an efficient way of getting a 50x trade done in an x by x market. There’s a reason people who need to hedge in an illiquid market do so ‘discreetly’, and it’s not that they can’t or aren’t allowed to place it all as one big limit order. Again you’re basically assuming away or ignoring the actual incentives and choices made by actual actors in the context (illiquid markets) that was under discussion.
The claim I thought you were replying to was simply that the meaning/benefits of transparency in liquid markets don’t nec. apply to illiquid ones. From what I see now, you’re not rebutting this claim so much as saying you wish for it not to be true. In particular, you don’t appear to dispute that hedge costs would be greater in illiquid markets with more price transparency; indeed, you’re saying you want them to be greater, for such-and-such reasons. Okay, but that’s a different discussion.
And if I’m evaluating that goal of yours I have to weigh the higher hedge costs, reporting costs, etc. you wish to impose, against whatever social benefits might come from imposing them. Far as I can tell you think the social benefits are that fewer deals/risk would be transferred, which would (somehow) ‘reduce the likelihood of crisis’.
If I don’t think it will, where does that leave me? I guess the underlying problem here is that we have different diagnoses of what caused the crisis.
best,
There’s a reason people who need to hedge in an illiquid market do so ‘discreetly’
What I was hoping was that you would explain why it is strategically better to trade sequentially rather than to place a large limit order and hold to your limit price. If it is true that this is a better strategy, then you should be able to explain why it is a better strategy instead of just telling me that you know it is.
What I’m guessing is that the advantage of trading sequentially is just a way to exploit the asymmetric information inherent in non-transparent, but presumably you have a more innocuous explanation.
The claim I thought you were replying to was simply that the meaning/benefits of transparency in liquid markets don’t nec. apply to illiquid ones.
Well, actually I think the burden of reply is on the proponents of non-transparency who need to explain what’s wrong with bond and option markets both of which trade very illiquid products in a transparent manner. Why isn’t anyone willing to address the question of what makes derivative markets so very different from other illiquid markets that a non-transparent structure is “necessary” otherwise there won’t be any financial transactions (as you asserted above)? It seems to me that trade in bonds and options trade functions reasonably well.
we have different diagnoses of what caused the crisis
Well, yes, I’d say that’s obviously true. But the real issue here is that proponents of non-transparent markets are just saying “Trust us. We know this is how it has to be. We know it’s the best way to balance social benefits and costs.” But at the same time, they guard their private information jealously to make sure that no one else has the ability to evaluate the claim (and the degree to which the claim arises out of the problem that “It is difficult to get a man to understand something, when his salary depends upon his not understanding it.”)
We trusted the claims of the financial industry up through 2007-08. The trust is gone now. Now is the time to use data and public information to prove your case, rather than simply asserting it and saying at the same time “Sorry, you can’t have access to the data, because non-transparency is ‘necessary.'”
What I was hoping was that you would explain why it is strategically better to trade sequentially rather than to place a large limit order and hold to your limit price.
I believe I already stated it: if all market actors (in particular, market-makers at large dealers) know that you have a giant bloc of risk that you need to hedge, one that hasn’t yet ground its way through the system, they know which way the price is likely to go, and thus can profit by front-running you. I wouldn’t think you would look kindly upon this or wish to encourage it, esp. as it’s precisely the kind of thing that dealers can and do take advantage of. If you don’t believe me, just ask an actual market-making trader. In particular, ask him if he’d like to know that so-and-so has a large position that he has decided to hedge via his product….
Also, I don’t consider it irrelevant to the argument to simply observe (as the Trioptima guys did) that when people do have a large bloc of risk they need to hedge, doing so ‘discreetly’ is how they, in fact, choose to do it. This revealed preference alone suggests that it is more efficient than placing the order all at once. Unless of course most market actors are wrong, and you are right that there’s no difference. But if so what is your opinion based on? Hunch?
What I’m guessing is that the advantage of trading sequentially is just a way to exploit the asymmetric information inherent in non-transparent, but presumably you have a more innocuous explanation.
It may surprise you to learn that I consider the preceding an already-innocuous explanation. The ‘asymmetric information’ in question is that I know the risks in my portfolio better than others do, and as such am better positioned to know what my needs are than others. So what? I consider ‘exploiting’ that knowledge to be totally innocuous and not sinister in the slightest. Why should others get to know my portfolio and its needs? What is wrong with me knowing more about my own property than others do? Absolutely. nothing.
need to explain what’s wrong with bond and option markets both of which trade very illiquid products in a transparent manner.
There are markets less liquid than the bond and equity-option markets. Actually, both look relatively liquid from where I sit. Maybe it’s all relative…
…what makes derivative markets so very different from other illiquid markets that a non-transparent structure is “necessary” otherwise there won’t be any financial transactions (as you asserted above)?
I didn’t say there wouldn’t be any financial transactions, I just agreed with Trioptima’s claim that they’d be more costly to hedge. And, you agreed too! Not only did you agree, but you said this was what you wanted. Not only that, but you wanted this precisely because you thought it would reduce the number of financial transactions, which you stated as your goal.
So, presumably, you can answer your own question; why do you think more transparency would lead to fewer financial transactions in such markets?
We trusted the claims of the financial industry up through 2007-08. The trust is gone now. Now is the time to use data and public information to prove your case, rather than simply asserting it and saying at the same time “Sorry, you can’t have access to the data, because non-transparency is ‘necessary.’”
I might understand this statement if you could make some logical connection between ‘lack of price transparency’ and the cause(s) of the crisis.
More generally, to be clear about something, I’m not a ‘proponent of non-transparency’. I’m a skeptic of government-imposed transparency for the sake of transparency. In particular, I’m prepared to believe that more transparency can and should be required in this or that context, if I can be convinced of the social benefits of doing so. But, I just don’t see them here.
You’re saying in response that it doesn’t matter, that (‘because they lost our trust’ or whatever) the burden is on people who dispute the need to require more transparency. I disagree. Absent a valid or compelling reason otherwise, the details of some private transaction entered into between A and B is none of C’s business. Again, there are times where there are compelling reasons to the contrary – but I am not seeing any here.
best
I didn’t say there wouldn’t be any financial transactions, I just agreed with Trioptima’s claim that they’d be more costly to hedge.
But if it’s just a change in the costs of hedging, why object to it? Why not gather data about a different derivatives trading environment so that we can all be better informed about derivatives trading?
In particular, I’m prepared to believe that more transparency can and should be required in this or that context, if I can be convinced of the social benefits of doing so.
So you’re willing to “be convinced of the social benefits” of transparency, but you require some kind of evidence. The catch-22 here is that the data isn’t public, so there’s no way for someone like me to put together a solid case.
What’s so objectionable here is that by avoiding transparency and keeping the data away from the public, the financial industry is also preventing the public from interpreting the data and thus limiting the discussion of the data.
In short, in the absence of transparency it’s difficult, if not impossible for outsiders, to gather enough data to make the demonstration of the value of transparency that you are seeking.
That’s why I brought the issue of the loss of trust in the financial industry (which is a much broader point than that of price transparency alone). If the financial industry wants to trade in dark markets, then we have to trust it to regulate those markets from the point of view of systemic stability itself. But the financial industry has already demonstrated that it is incapable of managing systemic risk in dark markets (like the shadow banking system). So that’s what I mean when I say the trust is gone.
As long as government/regulators/the public are holding the bag when it comes to systemic stability then all three of those entities must also have access to transparent information about financial activities that have any potential for systemic risk (and the AIG debacle already demonstrated that OTC derivatives have that potential).
Not really expecting you to agree with me. But that’s my point of view.
But if it’s just a change in the costs of hedging, why object to it?
Heck, all else equal I’m inclined to object to the proposal ‘let’s make X more expensive’ for all X, unless I am given a good reason to do it. That’s precisely why, in my very first response to this post, I asked you what the purported ‘social benefits’ were. And, as I still don’t see them, I still object.
So you’re willing to “be convinced of the social benefits” of transparency, but you require some kind of evidence.
Yeah. Go figure…
The catch-22 here is that the data isn’t public, so there’s no way for someone like me to put together a solid case.
There’s not ‘no way’. You could make a convincing argument. It doesn’t have to involve or rely on this particular data you refer to. Or even, any data at all. It could be from first principles, logic, common sense, or a convincing toy model of the economy that we could both agree on. Actually, any of the latter would be more likely to convince me of the social benefits than some sort of data-driven, dry argument involving statistics and regressions (if only because data can always be manipulated or presented in misleading ways).
What’s so objectionable here is that by avoiding transparency and keeping the data away from the public
Again, let me just clarify that you are using the phrase ‘keeping data away from the public’ to denote ‘not broadcasting the details of private contracts one has entered into’. For example, in the exact same sense, you yourself are ‘keeping data away from the public‘ by not posting PDFs of your mortgage documents/rental agreement, and/or employment contract with salary info, to this blog.
And again, I still think this is a weird way to talk, and am led back to: you want data on the private financial transactions of other people? Maybe, but where’s the argument that you merit it?
In short, in the absence of transparency it’s difficult, if not impossible for outsiders, to gather enough data to make the demonstration of the value of transparency that you are seeking.
Even more fundamentally, it’s difficult, if not impossible, for you to establish that not having such data has caused you or anyone else any problems. But all this makes me wonder is why you want or think you need it so much in the first place…
But the financial industry has already demonstrated that it is incapable of managing systemic risk in dark markets (like the shadow banking system). So that’s what I mean when I say the trust is gone.
Understood. But what does publishing price info on an OTC derivative have to do with any of that? How would it help avert collapses in systemic risk exactly? By what mechanism? Again: what are the ‘social benefits’ here, exactly?
As long as government/regulators/the public are holding the bag when it comes to systemic stability then all three of those entities must also have access to transparent information about financial activities that have any potential for systemic risk (and the AIG debacle already demonstrated that OTC derivatives have that potential).
The government, the regulators, and the public need not have held the bag for AIG (i.e. Goldman). The fact that they/we did was only due to a wrongheaded decision of specific actors within our government. It was neither a foregone conclusion nor an inevitable or predictable let alone even a logical outcome of any law or regulatory framework that was in place at the time.
To put it another way: it simply wasn’t the case that we stood to ‘hold the bag’ on AIG. We were inappropriately handed the bag by our esteemed representatives in government. Now, if you want to talk about how to prevent future AIG bailouts, my answer is simply, ‘Don’t do that anymore; don’t bail out the AIGs of the world; instead, follow the law’. Obviously that’s not what we did, but that is not my fault, nor is it the fault of markets having been ‘dark’. (Indeed I’m not sure there’s any reason to believe that had we known all about AIG’s contracts prior to then, the outcome would have been any different.)
Maybe this helps shed light on why I find this focus on ‘transparency’ and push to impose all sorts of further disclosure/publishing requirements to be misguided, if not a red herring entirely. ‘Better regulations’ don’t seem to have done us all that many favors in earlier episodes, and I see no reason to think that’s the answer now.
best
It could be from first principles, logic, common sense, or a convincing toy model of the economy that we could both agree on.
I’m working on that …
Which is why I find your comments/criticisms so useful!
the financial industry has already demonstrated that it is incapable of managing systemic risk in dark markets … But what does publishing price info on an OTC derivative have to do with any of that?
It’s the reason that (as happened in the 1930s) the burden of proof shifted to the financial institutions to be transparent or show the the value of non-transparency. (I understand that you don’t agree, but that’s my point of view.)
We were inappropriately handed the bag by our esteemed representatives in government.
Hmm. So do you think that AIG should have been put through bankruptcy?
I suspect that we can agree that the worst aspect of our financial system right now is the structural expectation of bailouts. In fact, in the absence of this structural expectation, I would almost certainly be less concerned about transparency. Unfortunately, it the world we have that needs to be regulated, not the world we want.
It’s the reason that (as happened in the 1930s) the burden of proof shifted to the financial institutions to be transparent or show the the value of non-transparency.
Not that I’m all that impressed with the regulations that came out of the 1930s either, but that’s a different discussion. Anyway, my question was more specific: what do you think the price transparency you say (perhaps correctly) should be presumptively required will actually accomplish? How exactly would it lead to prevention of crises, etc? i.e. Fill in the blanks: Once banks publish all their derivative prices, then ___ will happen, and that will prevent crises because ___.
Is the idea (for example) that with derivatives prices being highly public, then banks’ risk will be better managed? How so? Public pressure, Congressional nagging, what? I may be convinced that banks manage their risk horribly and irresponsibly, but that doesn’t automatically mean the public (or, who?) will do better.
So do you think that AIG should have been put through bankruptcy?
If they were insolvent, then of course that’s what should have happened. That’s what is supposed to happen to insolvent organizations. (Or a voluntary bondholder restructuring, or anything along those lines.) Isn’t it??
I am not a lawyer or expert but my understanding is that bankruptcy is how insolvent individuals/organizations may legally discharge their debts and that it is explicitly provided for in law and in the Constitution. ‘Bailouts’, in their various forms, are not. I think I would go so far as to say that I am against literally all ‘bailouts’.
Unfortunately, it the world we have that needs to be regulated, not the world we want.
I may agree that as a practical matter bailouts are going to happen, but any discussion of proposed regulations such as yours is a discussion of How Things Should Be. So I don’t see why I should accept the inevitability of bailouts when discussing proposed regulation changes; more to the point, if a set of regulations is tacitly premised on ubiquitous, anti-rule-of-law, corrupt, perhaps downright unlawful bailouts then this will enshrine such bailouts as business-as-usual, which presumably means bailouts are exactly what we’d get. I obviously disapprove and my favored policies reflect that.
best
what do you think the price transparency you say (perhaps correctly) should be presumptively required will actually accomplish?
As I said above, the real advantage of price transparency is that prices will move in response to hedging and this will cause financial institutions to internalize the liquidity costs of hedging. So the beneficiaries are financial institutions in the aggregate, in other words the financial system.
1. One problem is that it’s not necessarily ‘financial institutions’ we’re talking about here though. Financial institutions are in the middle; the end buyer of the risk may be e.g. a corporate CFO, or a public-employees pension fund, and so that’s whose hedge cost you are increasing. It seems like you think all financial actors are financial institutions, again that would be a puzzling world and one would wonder why any transactions ever took place 🙂
2. I see ‘the hedge costs one pays when putting on risk’ as a completely different animal from ‘the liquidity costs one pays in a liquidity crisis’. Increasing the former 1-5% doesn’t really have anything to do with preventing or reducing the latter. Or if it does, how?
You mean, just by making sure ‘fewer deals get done’? So at the end of the day, the social benefits you claim for this proposal are that they would reduce/suppress financial activity and the end placement of risk?
I am not sure I see that as a net-benefit…
best
I see ‘the hedge costs one pays when putting on risk’ as a completely different animal from ‘the liquidity costs one pays in a liquidity crisis’.
That’s the problem then. There are clearly liquidity costs to every purchase on illiquid market. You appear to want to assume those liquidity costs away, whereas I think they should be made as obvious as possible.
And yes the whole point of forcing financial institutions to internalize the costs of hedging is to pass those costs on to the end-user — creating a more accurate pricing structure throughout the financial system. I think it depends on the specific transaction we are discussing whether it is the end-user or the financial institution that faces the direct cost of hedging. I think you will agree that financial institutions hedge more (when this includes selling off end user risk that the financial institution took on) than end-users do.
The problem is that the current deals being done do not incorporate the liquidity costs of those deals, instead perfect liquidity is assumed. When the full cost structure of the deal is incorporated into it the financial system will be more stable.
That’s the problem then. There are clearly liquidity costs to every purchase on illiquid market. You appear to want to assume those liquidity costs away, […]
Not at all, and I’m really saying something different here. Let’s say that your policies lead to hedge costs of 1% of notional. (I made up this number). Then, oh, 4.5 years down the road, there’s a liquidity crisis, markdowns and collateral calls everywhere, and the person has to sell at a fire-sale price, which turns out to be 20 points below fair value (however defined).
I’m just saying the 1% upfront now has no connection to or salutary/mitigating effect on the 20% destined to be the fire-sale haircut. To put it another way, paying 1% now isn’t ‘internalizing the liquidity costs’ at all. It has no connection to the liquidity costs because it’s not based on the liquidity costs, it’s based on the compliance and friction costs created by your new regulation. The only thing your new cost and ‘contingent fire-sale liquidity costs’ really have in common is the same sign.
All of which is to say, basically, I don’t think adding this cost into the markets can, even in principle, have the positive effect you claim for it.
The problem is that the current deals being done do not incorporate the liquidity costs of those deals, instead perfect liquidity is assumed.
I don’t think this is true. Actually, I know it’s not.
best
When it comes to liquidity we’re really talking past each other. The liquidity cost that I’m talking about is just the effect of your transaction on the current market price, which you posit to be 1%. You internalize that cost by paying the lower price. The internalization has already taken place at the time of the transaction.
The only relationship between this and a future 20% loss in a crisis, is that you’re used to thinking that if you have to sell you’ll see losses. And if you have to sell in large quantities, you’ll see large losses — in fact the 20% losses in the “liquidity crisis” may well be explained in no small part by a change in the quantity sold.
It precisely because you are arguing that there are no liquidity costs in normal “illiquid” markets (which I perceive as a direct contradiction), that I claim that you are assuming perfect liquidity. If you weren’t assuming perfect liquidity, you would recognize that the price effect caused by a buying on an illiquid market is — or at least includes — a liquidity cost. (On a liquid market, your transaction wouldn’t have a price effect.)
There is always some nonzero liquidity cost and we are not in disagreement about that. But here, you are trying to make it larger than it currently is – which is to say, larger than it needs to be. No?
You have an argument for this which involves saying they should pay greater liquidity costs now to reduce/mitigate the [liquidity costs in a crisis, or something]. I’m just saying the former has nothing to do with the latter, and, so, it will not do that. Am I wrong?
thanks,
you are trying to make it larger than it currently is – which is to say, larger than it needs to be
The question is whether the existing market structure hides the liquidity costs of a purchase (preventing them from being fully reflected in the price) and , if so, to what degree this takes place. As an example, when John Paulson succeeds in buying protection on the CDS market, and the subsequent “asset” is then packaged into a CDO which is marketed to investors as synthetic CDO, but not as a bespoke CDO (that is, it’s not clear whether each buyer of the CDO understood that it was buying not 5% of the tranches sold, but 50%), does this lack of transparency affect the price at which CDS protection is sold.
If you can agree that such structures are capable of affecting the prices of the CDS traded, then we can ask whether the existing market reflects the liquidity costs or does not. If it does not, which I think is at least a strong possibility, then making that cost larger is not making it “larger than it needs to be”, but a correction of mispricing that is taking place.
In other words, your analysis assumes that the existing market prices liquidity correctly. This is an assertion that remains to be demonstrated.
Actually I doubt that any market prices, in anything, could ever be proven to be ‘correct’. I presume mispricing is the rule, and the only question is by how much. That said I’m inclined to place the burden of proof on the person claiming X is being (significantly) ‘mispriced’ to an extent requiring government action (which is what you are saying) to actually demonstrate as much. Given that you have asserted that “the current deals being done do not incorporate the liquidity costs of those deals”, an assertion I basically know to be untrue, it’s fair of me to wonder what you’re basing your belief on. At least, while you may have some valid reasons for believing that liquidity costs are being significantly mispriced, they are not in evidence here.
Let me backtrack slightly though, actually: I would agree with you that clearly, the pricing of CDOs back in 2005-2007 was often a mispricing in this sense, because they were priced off no-arbitrage to the CDS and index-tranche markets, which implicitly assumes perfect dynamic/replicating hedging (etc), artificially extending the assumption of liquidity to a product where it didn’t necessarily hold. But that was a specific product/scenario, and I don’t think it’s the case any longer – one has to establish this ‘mispricing’ on a case by case basis, and I don’t see that here.
The main point I wanted to make though was just that ‘the liquidity costs’ is not ever a constant number. For example, for a time, as far as I know the liquidity costs of CDOs back in 2006 may have actually been fairly low: it may have been possible to get in/out of them (relatively) easily, etc. The effect on the CDS market you mention is quite related to this – yes, the existence of CDOs affected the CDS market, similarly, the housing boom affected the granite market (I assume). But there’s nothing wrong with that by itself, i.e. those were genuine price changes and volumes in response to genuine demand. It’s just that *this changed* to the point where by 2008-2009 it no longer was liquid. That happens (right?). And the reason this is relevant is because the participants having paid an extra 1% (or whatever) in the ‘2006’ environment wouldn’t really have had anything to do with the liquidity costs that ended up being borne in the ‘2008-09’ sort of environment. Which was my point. The former will not stave off or even really mitigate the latter, so if that’s your argument for it, it falls flat.
In short, I’m not even disputing that mispricings can take place and cause distortions (though again, the burden’s on you to demonstrate this when/where it happens). I’m saying that your proposed solution doesn’t strike me as containing any mechanism for preventing this – i.e., it fails on its own terms as a solution to the problem cited, because I don’t think it is. Make sense?
P.S. Maybe I misunderstand but you don’t appear to be operating from the right definition of ‘bespoke CDO’. Bespoke just means, ‘not based on a market-standard index pool and its terms’. Most of what people refer to as CDOs, synthetic or not, were ‘bespoke’. In particular, not being bespoke has nothing to do with not knowing what tranche it’s exposed to (if that’s what you’re saying). Bespoke or not, the buyer should have known what tranche he held….
P.P.S. I want to say again thanks for these correspondences as I enjoy them & feel like I get a lot out of them. Feel free to let me know anytime I should cease & desist 🙂