Category Archives: Bonds

The changing collateral space

Izabella Kaminska and Cardiff Garcia from FT Alphaville, and David Murphy of Rivast, have written about how repo rates have detached from government yields.

Manmohan Singh from the IMF is out with a new paper that tries to explain what is driving collateral use.

Izzy has already done the heavy lifting on the paper and presentation, but the paper is quite good so here’s a quick overview from me. Basically, there are four main causes that affect the collateral market:

Although there are many new “entrants” and develpments in the collateral space, we highlight the four salient ones which in our view will have a non-trivial impact on the collateral dynamics. These are (i) central banks that remove good collateral from markets to their balance sheet where it is silo-ed; (ii) regulatory demands stemming from Basel III, Dodd Frank, EMIR etc that will entail building collateral buffers at banks, CCPs etc; (iii) collateral custodians who are striving to connect with the central security depositories (CSDs) to release collateral from silos; and (iv) net debt issuance from AAA/AA rated issuers.

First, central banks. Let’s use the SNB as an example:

Despite European Central Bank’s efforts to keep the ratio of good/bad collateral high in the EU financial markets, actions of SNB (and other central banks) are diluting this objective. Since the Swiss franc/€ peg in September 2011, the SNB balance sheet has quadrupled to about € 430 billion. The bulk of assets now comprise short tenor “core” € bonds (and lately UK gilts and US Treasuries). This reflects prudent asset-liability management at the SNB. However SNB’s bond purchases withdraw the best and most liquid collateral from the Eurozone; this reduces the collateral reuse rate since these bonds are silo-ed at SNB and not pledged in the financial markets. Silo-ed collateral has zero velocity by definition. Market sources indicate that lately long tenor € bonds also being purchased by SNB since there is limited supply of short tenor € bonds. Market sources also indicate that other central banks have also been buying core € bonds as part of their reserve management strategy; this again silos good € bonds and dilutes ECB’s efforts. Not surprisingly ECB’s next steps have been to expand collateral eligibility that includes lowering the asset based securities threshold, and relaxing the foreign-exchange collateral requirement (i.e., non-€ collateral is eligible).

It is the same with other QE-conducting central banks. They take good collateral out of the system and store it in “silos” so it can’t be used. The Fed is housing around $2.5 trillion of “good collateral” (largely U.S. Treasuries and MBS) — and, as Singh says, “at this rate Fed could silo over $1 trillion additional good collateral in 2013″. The Bank of England has around £375 billion gilts on its balance sheet; and the Bank of Japan is expected to buy about ¥15 trillion ($180 billion) of JGBs.

The second theme is regulation. I’ll let Singh explain:

Regulatory demands stemming from Basel III, Dodd Frank etc., are expected to silo $2-$4 trillion of collateral. Higher liquidity coverage ratio(s) (or LCRs) at banks, along with collateral needs for CCPs (and especially non-cleared OTC derivatives) are some of the key regulatory changes that will impact collateral markets. Recent proposals include provisions for sub AAA/AA assets to qualify for LCR. Similarly, eligible collateral at some CCPs already includes commodities (e.g., gold), corporate bonds (CME), equities, MBS (e.g., LCH Swapclear), renminbi (CME) etc. Some CCPs (especially at the request of buy side) are now seeking alternatives to title transfer that will allow clients to pledge securities as margin collateral. These safety buffers will silo the associated collateral and drain collateral from the financial markets. However, the regulatory proposals are staggered with changes that will be phased-in over the next five years.

Next up is what Singh rather boringly refers to as ”collateral custodians/depositories’, but it is very important. You might even spot a favorite or two:

In summer 2011, Eurozone had about €14 trillion in eligible collateral but much of it is either (i) locked in “depositories” and thus not easily accessible for cross border use; or, (ii) the asymmetry of demand for collateral in the ‘peripheral’ countries relative to the ‘core’ countries was at odds with the sizable availability of eligible collateral. Recently however, Euroclear and Clearstream (the key hubs for Eurozone collateral) are working with the local CSDs (or national/central security depositories) to alleviate collateral constraints. The interconnections to the CSDs will be via the Target 2 Securities (T2S) system that will provide a single pan European platform for securities settlement in central bank money.

Preliminary estimates suggest that perhaps up to €1- €1.5 trillion of AAA/AA quality collateral may be unlocked in the medium term via efforts of custodians to optimize collateral and build a collateral highway or global liquidity hub. However, the internal “plumbing” (i.e., operations, workflows, technology, staff etc) that is required to process and manage trillions of collateral balances needs to be smooth.16 Even if this collateral does not reach “large banks/markets”, it allows the collateral to leave “CSD silos”, improve efficiency, and enhance accounting debt and credits and reduce the burden on markets to provide collateral for LCR or CCP related regulatory buffers.

Okay, so not all bad, although ‘solved’ might be an over statement, as evident in footnote 16:

16. Every institution or market is different; there is a lot of friction in the pipes. Even though legally collateral is allowed to be re-used but if a counterparty along the collateral chain hasn’t built the system to do anything with it, the collateral gets “stuck” in the plumbing. The frictions in aggregate can be quite sizeable and may be another reason why the theoretical balances may not add up mathematically.

Finally, we have new net issuance of debt:

Assuming AAA/AA countries have GDP of around $25 trillion and with a deficit of around 4-5 percent, they have supplied (on average) about $1 trillion of new (net) debt—sovereign and corporate–every year, with latest data on the lower side. 20 Database and market contacts suggest that on average about 30%-40% of AAA/AA collateral inventory reaches markets via custodians for re-use (on behalf of reserve managers, SWF, pension, insurers etc); however, much of the inventory stays with buy and hold investors. So if Debt/GDP remains on trend in developed countries (i.e., the ratio does not increase sizeably), new debt stemming from the “numerator” may provide up to $300-$400 billion per year to the markets assuming counterparty risk especially with European banks does not elevate. Another 5%-10% of new inventory (including equities) may come via hedge funds. With a collateral reuse rate of about 2.5 (and assuming this does not decrease due to the various silo(s) in the “new” collateral space) this may alleviate collateral shortage by about $800 billion to $1.2 trillion per year.

And here’s how it all looks graphically:

Screen Shot 2013 01 30 at 1.24.23 AM The changing collateral space

And so Singh suggest a couple of policies. Central banks could, for example, “rent” out their good collateral (and their balance sheet) as this will increase collateral velocity as the good/bad collateral ratio will rise.

In summary, the tone over the medium term is somewhat positive, but there is in my opinion a very big *if*:

Collateral scarcity is distinct from collateral shortage. 31 Annual collateral “flows” (e.g., via new net debt issuance) may seem less relative to desired collateral “stocks” (e.g., for regulatory buffers) but that may not be the case over a five year cycle. Several competing developments have been in motion in recent years with positive and negative implications for the collateral market s: ECB’s LTROs (+), Fed and SNB actions (-); new net debt issuance (+); collateral velocity (-). In the near future, sustaining collateral velocity (i.e., collateral re-use by markets) will be important since there is an unfortunate constellation—from central bank actions and regulatory proposals—that will silo good collateral. Fine-tuning some regulatory demands will alleviate collateral scarcity in the near term. So will actions from some central banks that have vested interest to provide good collateral in the markets. In the medium term, assuming no major dislocation in financial markets, central banks are likely to unwind sizable good collateral from their balance sheet and alleviate shortages (if any).

Target2 – a quick note

David Keohane over at FT Alphaville highlighted this little chart from JP Morgan’s F&L team on Monday:

Target2 balances JPM Target2   a quick note

So we’re seeing a reversal of previous trends, albeit a small one (and I’m not to say what the reasons are – possibilities include ECB actions, money market funds entering Europe, SNB, economic growth, investment opportunities – who knows).

So why is this important? Because it is a very good window into intra-Eurozone capital flows. Forget for a moment the politics of it (including the transfer debate). The problem before and now is that the private capital flows from German banks were replaced by public lending via Target2/central banks. I thought Martin Wolf did a very good job summing it up a week ago:

First, Germany has accumulated net claims on the rest of the world – and on other members of the eurozone – not because of internal central bank accounting, but because it has large current account surpluses. Germans have been running two businesses: exporting goods, at which they are excellent, and importing financial claims, at which they are not. In brief, Germany’s surpluses have exposed Germans to financial risk. But balances inside the eurosystem are not a good indicator of that risk. They have exploded, argues the paper, because of speculative financial flows, not current account imbalances.

The problem isn’t that central banks can clear money transactions – that’s essential in a monetary union. The problem is that it allows money to flow very easily from accounts in the periphery to the core, pushing yields into negative in the core, creating a broken transmission mechanism for monetary policy, and leaving the periphery with a capital shortfall.

Normally in a monetary union, the central bank can set interest rates and that’s that. In the Eurozone, however, that’s not the case. Consider Germany: Banks are flush with capital, everyone wants Bunds, and banks are willing to lend at low rates. Now consider Spain: the banks are insolvent, nobody wants Spanish government paper, and businesses can’t get loans. Two very different situations with one monetary policy. Poor Draghi.

So when the ECB lowers the interest rate it doesn’t really matter because it isn’t transmitted through to the periphery. Add to that the fact that not that many use unsecured lending any more, which is what central banks have a monopoly on, and it gets interesting.

None of this is new, of course. FT Alphaville has written a lot about it (me too), but I thought it worthwhile to write a little. Target2 is an enabler of all this – not the problem. It does, however, provide a nice window into what is happening in Europe.

So is the reversal of capital flows a good thing? Yes. Let’s see how long it lasts…

Yes, Sweden’s floating currency helped them

I thought this was very persuasive from Brad DeLong on the Swedish economy, a lecture on how not to discuss economics. Something for the Danish politicians to think about:

Listening to Sweden’s Finance Minister Anders Borg on Charlie Rose was, I found, quite frightening.

At least where I sit, the biggest piece of Sweden’s response to the downturn was an aggressively expansionary monetary policy that lowered the value of the kroner by 15% just as the downturn hit and switched a lot of European demand into Sweden:

 Yes, Swedens floating currency helped them

In a small open economy like Sweden, even small declines in the value of the currency can generate large increases in aggregate demand. That was what kept the rise in the Swedish unemployment rate at a mere 2% points, compared to 5% points here in the United States:

 Yes, Swedens floating currency helped them

But Borg was, instead, parroting the Alesina-Ardagna line akrateros, claiming that tight fiscal policy had worked wonders in Sweden–and that the reason that Germany had a smaller downturn than the U.S. was that Germany had undertaken less expansionary fiscal stimulus.

First of all, that claim simply is not true: the differences in discretionary fiscal stimulus undertaken by Germany and the U.S. were trivial:

 Yes, Swedens floating currency helped them

Second, if it were to be true, it would have been because less fiscal stimulus in Germany would have boosted confidence, generated greater declines in interest rates, and so boosted private investment spending. But the declines in interest rates in Germany and the U.S. during this crisis have been equivalent: both are creditworthy governments benefitting from a flight to quality:

 Yes, Swedens floating currency helped them

 Yes, Swedens floating currency helped them

Why oh why can’t we have better European finance ministers?

Target2, flow of funds, and liquidity providing banks

The debate is still going on whether Target2 is a big problem or nothing more than an aberration. I tend to be in the first camp. What I see as the problem is that, because of Target2 imbalances, monetary policy is less effective and limits the political flexibility. Here are my reasons:

1) The excess liquidity that flows into mostly German banks make monetary policy harder, because it distorts the usual interest rate mechanism by which the ECB can influence rates, thus leaving repo-rates and the liquidity channel to target.

2) People say that it is only a problem if the eurozone breaks, at which point we’ll have more problems than Target2 imbalances. Not so. The mere existence of Target2 imbalances make a break-up much less likely, because the bigger the liabilities, the more the core would stand to lose in a break-up.

3) It provides an interesting view of the actual money flows in the eurozone.

Case in point of the last one is here. The ECB’s Eurosystem Oversight Report. Buried on page 49 is a graph of how important banks function as ‘liquidity providers’ within communities of the eurosystem. In the ECB’s words:

A clear view of the network of payments is useful to overseers for various reasons. It helps to identify banks which are “central” from a liquidity circulation point of view. It can reveal possible dependencies between banks in terms of liquidity provision. Moreover, observing the network over time can reveal trends or anomalies. If these lead to stress events, the anomalies may be used as early warning indicators.

Therefore, liquidity can be viewed as flowing between the big consolidated banks and, from these, to the smaller banks.

Graphed that looks like this:

Picture 3 262x300 Target2, flow of funds, and liquidity providing banksPicture 4 233x300 Target2, flow of funds, and liquidity providing banksPicture 5 269x300 Target2, flow of funds, and liquidity providing banks

If I understand it right, a few large banks are thus — now more than ever — too big to fail, because an important part of the payment system runs through the large banks.

Target2 gives insight to the flow of funds in the eurozone. If I read that right, this is a strong argument against a break-up of the eurozone, because the technical consequences would be large, in addition to the assets the Bundesbank would lose, as a Greek exit (to pick randomly) would probably mean a default on those obligations.

It is going to be interesting to see the Target2 flows from Spain in the coming months. If we see a continued outflow of funds, and thus a rise in liabilities, it could, ironically, reduce the risk of a break-up. What a big-bank bankruptcy would do is left with no answer..

Denmark’s CDS… again

Denmark is a safe haven. Surely, we can agree on that. Just like the other safe havens, we’re able to borrow for a long time at a very low interest rate. Only Switzerland is able to borrow at a lower interest rate than Denmark (and Sweden in the 30-year bond). Here’s Denmark yield curve (Source: Bloomberg):

Denmark yield curve Denmarks CDS... again

Very low interest rates, indeed. Now, are they low in comparison to whom we usually compare us to? Yes (source: Bloomberg):

Yields spreads comp1 Denmarks CDS... again

Now, here is the puzzle. Why is there an apparent mismatch between bond yields — where we do just fine — and credit default swaps? Here are the CDSs on sovereign debt for the same countries (source: Bloomberg):

CDS1 Denmarks CDS... againI’m using 5-year CDS, by the way, because it is (as far as I know) the most used and liquid. Other maturities yield the same results (according to Bloomberg).

Now, why is it relatively more expensive to insure against a Danish default than other sovereigns, especially since we are able to borrow at cheaper rates than them? It doesn’t seem to me to make much sense.

I’d like to note a couple of possible reasons, though. 1) The Danish financial sector is huge comparable to the size of the whole economy. 2) Danish institutionals (pension funds, insurance companies, etc.) buy lots of bonds, but are not buying CDSs. 3) The CDS market is a hedge on Denmark. 4) The Danish banking sector is not that well, at least after Amagerbanken went bankrupt. 5) The excess liquidity created by central banks has gone into bonds (although this seems less likely as Danish interests rates have been fairly stable over time).

I don’t really get it, though, because we do have the option of breaking our peg with the euro, and there seems to be big arbitrage possibilities. Maybe it’s just one of those things where we haven’t really figured out how big a risk we have with our financial institutions (or our housing market, exports, or…).

Fiscal policy in a depressed economy

Brad DeLong and Larry Summers has a new paper out titled ‘Fiscal Policy in a Depressed Economy‘. It’s full of greek letters, but it’s conclusion is quite clear: Short term fiscal stimulus works and is arithmetically worth doing, even when taking account of the debt issue. I won’t quote the paper (you should go read it), but I’ll post three graphs, which I think sums up their argument nicely.

First, we have the real interest rate at which the government can borrow. What we see is that the real interest rate is negative, thus ensuring that if the investment the government undertakes has a positive return, it’s worth doing:

Picture 41 Fiscal policy in a depressed economyNext, the fiscal multiplier, i.e., if the government spends $1 how much of that dollar will get spent. Think of it as if the government builds a bridge. Then the government pays the workers directly, but then the workers use their wages to buy goods and services, etc.

Picture 73 Fiscal policy in a depressed economyFinally, we have the U.S. employment-to-population and unemployment ratio:

Picture 5 Fiscal policy in a depressed economyThis is essential. In “normal times” (where short-term rates aren’t up against the zero lower bound), fiscal stimulus can crowd out investment (by raising interest rates). This is not the case in a depressed economy, which is in a liquidity trap. As a matter of arithmetics, short-term and medium-term stimulus is worth doing, because the multiplier is substantial and a sell-off in government bonds is unlikely. Thus (part of) the conclusion:

Our analysis simply demonstrates that additional fiscal stimulus, maintained during a period when economic circumstances are such that multiplier and hysteresis effects are significant and then removed, will ease rather than exacerbate the government’s long run budget constraint.

I’m not doing the paper justice as I haven’t talked about monetary policy, reservations, limitations, etc. — so go read it.

Financial stability and fiscal instability

So, the ECB is out with its monthly bulletin. As always, it’s filled with nice graphs and data.

I want to make a few very simple points. The first is that since the first LTRO, volatility has fallen and markets have risen (for whatever reason). Presented graphically we have stock markets and implied volatility:

Picture 8 214x300 Financial stability and fiscal instabilityPicture 71 192x300 Financial stability and fiscal instabilitySo far so good. It’s hard to make the right decisions when meetings are interrupted by messages of flash crashes, panics, and updates that your pension fund had allocated all your funds to equities. So equity vol has come down. Are we safe? Let’s have a look at government bond yields spreads to bunds:

Picture 9 Financial stability and fiscal instabilityHm. Not quite there yet, but it’s better than in late-2011.

What about the European people? Is austerity the magic cure for all of our problems? Nah, I didn’t think so:

Picture 4 Financial stability and fiscal instabilityIt is really hard to help the unemployed with no growth, and there is correlation between growth and employment. It doesn’t help looking at the scariest chart (youth unemployment). What do I mean? Well, I mean the ECB’s outlook might look a little rosy:

Picture 81 Financial stability and fiscal instability

Finally, how about those banks? Did the LTRO help? Well, we can say for sure that the higher the funding costs (and thus the stress of the bank, presumably), the higher the amount of LTRO funds they took:

Picture 72 Financial stability and fiscal instabilityThis is really important, as the ECB is — as far as I know — the only organization that have all the numbers. We can thus say, with some statistical certainty (see the regression) that the higher the stress of the bank, and the harder it has to access the funding market, the more money it took in the LTRO. Also, CDSs have fallen. Funding is easy now.

Democratic Governance

As previously mentioned, the real story about the European crisis is one of lacking governance structures, and the struggle to create a new durable fiscal compact in the midst of the crisis. Politics as a whole becomes infinitely more difficult when confronted with the resource shortages that can prevent both coalition building in parliaments and patronage-based election strategies. Francis Fukuyama has a brilliant piece in his newly revived blog at the American Interest that touches on the problem of “good governance”:

Conversely, I would argue that the quality of governance in the US tends to be low precisely because of a continuing tradition of Jacksonian populism. Americans with their democratic roots generally do not trust elite bureaucrats to the extent that the French, Germans, British, or Japanese have in years past. This distrust leads to micromanagement by Congress through proliferating rules and complex, self-contradictory legislative mandates which make poor quality governance a self-fulfilling prophecy. The US is thus caught in a low-level equilibrium trap, in which a hobbled bureaucracy validates everyone’s view that the government can’t do anything competently. The origins of this, as Martin Shefter pointed out many years ago, is due to the fact that democracy preceded bureaucratic consolidation in contrast to European democracies that arose out of aristocratic regimes.

This quote is absolutely crucial, and serves to underline the incredible complexity that lies in connecting nations with different national governance systems, different varieties of capitalism and different expectations for the transnational project. Putnams theory of two-level games illustrates the added complexity that comes with achieving domestic policy goals while attempting to build diplomatic ties with international partners. Add a supernational court and a “technocratic” commission and this job does not become easier.

The Danish discourse is often vaguely neoconservative in its imagery, with the European community now “helping” Greece not just in an economic sense but in a cultural and democratic one. Clearly this is a rather vague version of Huntington’s old theory of authoritarian transitions (Fukuyama describes this here). How much of Greece’s supposed acceptance of governmental change (but not reforms) is due to framing effects, and how much stems from true alignment with the policy goals of incumbent technocrats? The drama continues, and surely one of the key issues will be whether Greece chooses to impose its rather draconian bond laws on private investors. If this is the case, this is one blow against liberalism and free markets that will be very hard to condone as the swan song of the “old Southern Europe”.

Regardless, Greece is absolutely insolvent and there is no reason to beat around this conclusion. Felix Salmon has this succint summary of the situation, with Germany essentially setting the stage for an imminent Greek default. This will not be a smooth ride, but it is better to get it done as quickly as possible, for the populations in all countries. And with the ECB having prepared the Eurozone for this event, i see no reason that Greece will not have defaulted within a very short (a month or so) timespan, regardless of what politicians and the media will choose to call the event. And as Simon has pointed out, the handling of credit events in terms of CDS exposures will be as important as the Greek bond laws in establishing the proper financial regulatory framework that will prevail after the default.

European Interest Rates, Graphed

The IMF has an update to their Fiscal Monitor, which has a nice graph that shows just how skewed European yields are:

Picture 11 European Interest Rates, Graphed

Having Fun With the ECB Monthly Bulletin

The monthly bulletin from the ECB is out. That means lots of cool graphs and data. Since I’m studying for my Statistics exam, I thought I’d pass along a few of the more fun technical figures.

First up in chart A, we have the bidding behaviour of banks for the LTRO (as a percentage of total assets), on the x-axis, and the amount of long-term debt securities maturing in 2012-14 on the y-axis. What this show is that there is a positive relationship between long-term debt that needs to be rolled over in the next three years and the use of the LTRO for banks. The more debt that matures and needs to be rolled over in the next three years, the more the banks wanted to tap the LTRO. Chart B shows this relationship by plotting the residual maturity of outstanding debt against the same bidding behaviour. If a bank is well-funded for the next three years, it’s less likely to use the LTRO (or use less). That’s what these two charts show. For the Statistics interested, remember that we only have one sample to go from; the bidding is secret (we don’t know who’s hiding behind the dots); it’s not random (only troubled banks use it); and the funding structure for some banks might make it necessary to use the LTRO because of market turmoil (the interbank market is essentially frozen).

Picture 4 Having Fun With the ECB Monthly BulletinThen, we have a chart of deposits held at euro MFIs (Monetary Financial Institution) by sector. This graph goes a long way to show why some banks have liquidity and funding problems, and why the market is so much more volatile than before 2008:

Picture 6 Having Fun With the ECB Monthly BulletinAnd then volatility. Chart A shows a few different measures (one for the US, two for the euro area). What they show is that we’ve had some very volatility times. We’ve had one event (2008) with a standard deviation of three (probability of that is 0.3 %) — and it actually has two peaks in one measure. We’ve also had two events that fall almost two standard deviations away from the mean (2001 and now). Probability of that is around 5 %. Tail risk, indeed. Not hard to gather why it’s difficult to fund oneself if we have 2 two-standard deviation events and one three standard deviation event in ten years (including three one-standard deviation events). Chart B shows the financial market stress for each market (FX, equity, and banking)*:

Picture 71 Having Fun With the ECB Monthly Bulletin

Finally, to show how the debt issuance has fallen (as a result of volatility, the financial crisis, recession, lack of demand, etc.), this graph pretty much says it all. Note how general government issuance has fallen (due to austerity) since 2009:

Picture 8 Having Fun With the ECB Monthly Bulletin

Could it be that we might need some more safe haven assets (H/T DeLong for the last one)? Anyway, nice graphs, statistics, data, and lots of math to drive one away from the waaaah lovely books of Statistics. See the appendix of the bulletin for the formulas, if you’re really bored interested.

*The sources for the volatility measures are:

1) Bekaert, G., Hoerova, M. and Lo Duca, M., “Risk, uncertainty and monetary policy”, NBER Working Paper Series, No 16397, National Bureau of Economic Research, 2010. This measure also provides an estimate of risk aversion.

2) Greiber, C. and Lemke, W., “Money demand and macroeconomic uncertainty”, Discussion Paper Series 1, No 26, Deutsche Bundesbank, 2005.

3) van Roye, B., “Financial stress and economic activity in Germany and the Euro Area”, Kiel Working Papers, No 1743, Kiel, Institute for the World Economy, 2011.

4) The European Central Bank