Book
Share
Chapter

4. Collateral and Monetary Policy

Author(s):
Manmohan Singh
Published Date:
October 2016
Share
  • ShareShare
Show Summary Details

This chapter deals with the relative prices of money and collateral as both contribute towards financial lubrication in the markets. With some central banks now a major player in the collateral markets, the larger the quantitative easing efforts, the longer these central banks will impact the collateral market and associated repo rate. This may have monetary-policy and financial-stability implications, since the repo rates map the financial landscape that straddles the bank–non-bank nexus.

Introduction

Collateral is not integrated within the money or monetary-policy textbooks. Undergraduate macroeconomic textbooks still use the IS/LM (standing for investment, saving, liquidity preference and money supply) model as a construct to demonstrate the relationship between interest rates and real output in the goods-and-services market and the money market. In this model (see Figure 4.1), the intersection of the IS and LM curves is where there is simultaneous equilibrium in both markets. In the figure, the horizontal axis represents output, or real GDP, and is labelled Y. The vertical axis represents the real interest rate, i. Since this is a non-dynamic model, there is a one-to-one relationship between the nominal interest rate and the real interest rate; therefore, variables such as money demand – variables that actually depend on the nominal interest rate – can equivalently be expressed as depending on the real interest rate. The point where these schedules intersect represents a short-run equilibrium in the real and monetary sectors. This equilibrium yields a unique combination of the nominal/real interest rate and real GDP. However, LM is only about money and does not include the sizable pledged-collateral market, which is equivalent to money when settling accounts.

Figure 4.1The IS/LM Model

The IS curve encompasses the investment and saving equilibrium; in other words, total spending equals total output or GDP of an economy. The output components are as consumption (C) + private investment (I) + government expenditure (G) + net exports (NX). Looked at simply, inward shifts in the IS curve (due to a contraction of the C + I + G + netX component, which decreases output to YB) can be neutralised by shifting LM out by lowering (nominal and thus real) interest rates to attain the initial level of output YA. Pioneering works on the financial accelerator highlight how endogenous shocks to credit markets can initiate cyclical effects to the real economy. See, for instance, the paper by Ben Bernanke, Mark Gertler and Simon Gilchrist (1996). Specifically, their paper highlights how shocks can result in flight to quality and thus higher cost for risky projects; this moves the IS curve inwards, since households and firms invest less. Subsequently, the authors show, the financial accelerator can amplify shocks stemming from collateral constraints to the economy.

However, financial collateral that substitutes for money is different from the general collateral modelled in the original research papers. From an overall financial-lubrication perspective that requires intraday debits and credits, the cross-border financial markets traditionally use “cash or cash equivalent” in lieu of posting margin (ie, money plus collateral). For illustration, since LM is strictly about central bank money, Figure 4.3 shows shifts in the pledged collateral market via the IS curve (with the intuition that pledged collateral can be viewed as an asset under I, or private investment)

Price of Money and Price of Collateral

In some countries, such as the US and the UK, the price of money and money market rates are not market-determined due to interest on excess reserves (IOER) offered only to depository institutions. This creates a wedge between banks and non-banks, and thus impacts other short-end rates. This has resulted in market segmentation and forms a wedge in the money market rates. Following the Lehman failure, the Fed introduced paying IOER to only the depository institutions. This was intended to place a “floor” (minimum bid) on short-term liquidity in the corridor system.

In the US, Fannie Mae, Freddie Mac, and other non-banks like money market funds (the “money arteries” of US plumbing) cannot access IOER (that pays 50 basis points (bp) at the time of writing), which only banks can receive. Therefore, Fannie and Freddie’s cash positions (and those of other home-loan banks and non-banks) have largely determined the federal funds rate, which trades below the IOER as these non-banks can only access the IOER via a bank. This wedge between IOER and the federal funds rate is important. This wedge has been higher (and over 10bp) since the Fed’s lift-off on December 16, 2015, when IOER moved from 25 to 50bp.

Now consider collateral or repo rates. Recall that the collateral or repo rate is that at which cash is lent against collateral for an agreed tenor. It is agreed on by the two parties at t0, or start of repo. Typically, collateral shortage lowers repo rates; collateral abundance increases repo rates. This rate is a proxy for collateralised transactions that underpin the financial plumbing between the dealer banks and non-banks.

From July 2012 through summer of 2013 in the eurozone, collateral/repo rates dipped below zero; since then they have been even lower (minus 40 bps and below) following the ECB’s QE program from March 2015. These include German, French, Dutch, Danish and Swiss repo rates in recent times.1 However, this is not the case with collateral/repo rates in the US. In theory, the price of “good collateral” should not vary across assets except due to technical factors that include “home” bias; liquidity, depth; relative fx rates; cheapest-to-deliver collateral; etc. The RRP (reverse repo program) in the US has been instrumental in maintaining a wedge between comparable repo rates in the US and those in the eurozone (Figure 4.2a and 4.2b). The Fed’s Operation Twist, which largely took place in the second half of 2012, also provided an extra dose of T-bills in 2012 to provide some lift to the general collateral financing (GCF) rates. Thus good collateral such as the US GCF rates remained in positive territory relative to good collateral in the eurozone that has been in negative territory. In the US, it remains to be seen if cash shifts from repo to bank deposits when overnight GCF rate turns negative. There is a big psychological barrier between explicitly paying for protection and accepting a lower return to get protection. However, with the Fed’s reverse-repo floor initially around 3 to 5bps, and at 25bps after the Fed’s lift-off, GCF rate is now unlikely to turn negative.

Figure 4.2aCollateral Rates in US

Source: ICAP, Bloomberg, DTCC and staff estimates.

Figure 4.2bSelected EU Countries

Source: ICAP, Bloomberg, DTCC and staff estimates.

Next, we look at the relationship between collateral and repo rates and monetary policy via the IS/LM framework.

Collateral and Monetary Policy via the IS/LM Framework

When collateral use drops, financial intermediation slows, with effects similar to the drying-up of interbank markets. The stock of collateral can decline as investors become more concerned about counterparty risk, making them less willing to lend securities resulting in idle collateral to sit in segregated accounts. It can also be affected by central-bank measures such as large-scale asset purchases, which drain good-quality collateral from the system, or a widening of the pool of collateral-eligible assets, which increases the pledgeability of these assets as collateral to the central banks.

Financial collateral does not have to be rated AAA/AA, but, as long as the securities (ie, debt or equity) are liquid, mark-to-market and part of a legal cross-border master agreement, they will be used as “cash equivalent”. Such pledged financial collateral is difficult to map but is a key component of financial plumbing. The collateral intermediation function is likely to become more important over time. In the short term, increased counterparty risks (as during 2007–8 in the US, and later in Europe) make secured funding more attractive. In the longer term, with more arm’s-length transactions in an increasingly globally integrated financial system, market participants are seeking the security of collateral to underpin a wider range of claims. New regulations are also likely to increase the demand for collateral-based operations (CGFS 2013).

The collapse in financial collateral since the Lehman crisis (by an estimated US$4–5 trillion) has significantly shifted the IS curve of IS/LM inwards, lowering output to YB and decreasing the real interest rate. In recent years, efforts in quantitative easing (QE) are shifting the LM curve to the right to compensate for this decline, until the LM curve will intersect with the IS curve at the initial output YA. The LM shift due to QE is sizable (and continuing) in the US and, along with the IS inward shift, real interest rates may be well below zero (but “optically”, due to distortions in the money rates that are above zero in nominal terms, we do not see subzero real rates). See Figure 4.3.

Figure 4.3Contraction in Pledged Collateral Market and IS/LM Shifts

A speech in 2012 by New York Fed president suggests that the US Fed’s QE actions may have lowered the nominal rates by an additional 150–200bp (Dudley 2012). So unadjusted real rates (ie, if the Fed balance sheet had remained the same at US$800 billion as of end-2007) may be much lower relative to the adjusted real rates (due to the expanded balance sheet via QE) that are officially announced and do not factor the rate cuts embedded within QE – the red and blue lines respectively in Figure 4.4. Now consider the three most recent tightening cycles since 1994 in the US that have averaged close to 400bp. With a 400bp tightening cycle, the new policy rate may anchor at 2% if it starts from minus 2%. Analytically, in Figure 4.3, LM shifts out until YA is reached at approximately minus 2%, and, with a 400bp tightening, the next policy rate cycle may stop at 2% (unless the balance sheet also returns to the 2007 size in tandem). Intuitively, QE resulted in lowering collateral velocity. From a financial lubrication angle (money plus collateral), lower collateral velocity contributes towards tightening. So the next monetary rate cycle (via interest rates) may not have to go the whole way.

Figure 4.4Real Interest Rates Via Taylor Rule with/without Fed’s Balance-Sheet Adjustment

Central Banks and Collateral Markets

Some central banks (such as the Fed, the Bank of England and recently the ECB and Bank of Japan) have become large repositories of good collateral as a result of their QE policies. But excess reserves at central banks are not the same thing as good collateral that circulates through the bank–non-bank nexus. As a result of this the bank–non-bank nexus over time has begun to give way to a new central-bank–non-bank nexus, which has weakened the market’s financial plumbing and increased shadow-banking “puts” to compensate for the lack of good collateral.

Although there are many variants and interpretations of “exit”, a key aspect is its impact on the part of the market where non-banks interact with the large dealer banks to determine the price of collateral (the repo rate, see Figure 4.5). Some central banks that have undertaken QE are now holding sizable amounts of high-quality liquid assets (or good collateral) on their balance sheets. Their proposals to unwind that inventory come in part to stem any shortage of good collateral. However, such proposals for unwinding will have implications for this part of the market in a way that may cause major adaptations to take place.

Figure 4.5Collateral and Financial Plumbing

While it’s true that, sooner or later, these central bank balance sheets will have to unwind – either voluntarily, when central banks release collateral and take in money, or involuntarily, as the securities held at central banks mature or roll off – unwinding will increase both the (money) interest rate and the (collateral) repo rate. In other words, collateral will get cheaper if the balance sheet unwinds and lets the securities reach the market; this will move the repo curve up. Money will get more expensive as rates increase; this will move the interest curve up.

In the US, the Fed has bought good collateral largely from non-banks, not banks (Carpenter et al 2013). This has increased bank deposits (that belong to non-banks via the QE money they got in lieu of collateral sold to the Fed). So the effect of QE-type efforts is to convert what had been good collateral into additional bank liabilities (ie, non-banks’ deposits at banks).

The Fed’s fixed-allotment reverse repo programme inaugurated on September 23, 2013, is the first official attempt to unwind part of its balance sheet. (In the initial days, the RRP use was capped at around US$1 billion per counterparty and was increased in a series of steps. On December 16, 2015, the RRP facility was elevated to about US$2 trillion and the maximum bid was kept at US$30 billion per counterparty that was prevailing since September 2014.) The success of this programme will be affected by allocation of balancesheet “space” between banks and non-banks amid a tighter regulatory environment. Non-banks’ “balance-sheet space” will be key to any unwinding of collateral. With Basel III regulations at the door (especially the leverage and LCR ratio), the banking system is likely to have limited appetite for increasing balance sheets.2 Reverse repos would actually reduce total bank balance sheets by the amount of reverse repo the Fed does with eligible non-banks such as money-market mutual funds (MMMFs), Fannie, Freddie, and select asset managers.

If we look at collateral chains, at one end there is the MMMF investor – the household and corporate wealth pool (the supplier of money). At the other end, after a couple of loops for transformation, and some haircuts and subordination for extra capital, lies the promise to pay made by the borrower – household (mortgage) or hedge fund. The Fed’s reverse repo relieves bank balance-sheet constraints but short-circuits the chain. The household and corporate wealth pool is better off: it gets a deposit alternative that is superior to anything available now. The borrower pool is worse off, as money will go directly to the Fed, and won’t be transformed into any lending to that pool. A simple illustration that gives the intuition: the Fed takes money via RRP directly from the money arteries of US plumbing (Fannie, Freddie, MMMFs etc). Thus, these non-banks withdraw money from the dealer/banks who will in turn return the securities (UST, MBS etc) back to securities lenders who were sec-lending to dealer/banks to augment returns. The dealer/banks will also give back securities to the hedge funds as these dealer/banks will not have the money to finance hedge funds via repo. So, demand for (and price of) such securities (including UST and MBS) falls. Thus the value of Fed assets falls – whether they sell them or carry out RRP with non-banks.

The truth is that excess reserves do not simply become “good collateral” as the central bank unwinds its balance sheet. This is primarily because collateral with these non-banks via reverse repos cannot be rehypothecated, or onwardly repledged, and thus will not contribute to financial lubrication. The reasoning for this is that two clearing banks (JPMorgan and Bank of New York) can support rehypothecation of securities in the triparty process only through what is called GCF (or general collateral finance), which is an interdealer triparty service (ie, banks) for members of the Government Securities Division of the Depository Trust & Clearing Corporation (DTCC). If you are not a GCF participant, you effectively have “read-only” access to its collateral (except in the case of default, for which it has a separate, more manual process to send securities to the customer custodian to facilitate a sale). Thus RRP with non-banks is not “reserve drainage” but an “accounting drainage” as securities still remain on the liability side of the Fed balance sheet - they are just renamed from “excess reserve of non-banks” to “RRP with non-banks”. Securities do not reach the market! This puts a lid on releasing collateral velocity.

Only banks are able to rehypothecate collateral received via reverse repo (and increase collateral velocity) if (i) they have balancesheet space, and (ii) to the extent that collateral from RRP substitutes for otherwise available securities that the dealer bank was pledging in triparty, the RRP can free up securities to go into the bilateral market. Thus, the new central-bank–non-bank nexus is good for non-banks, since the collateral counterparty is the central bank. But it is also an extension of the Fed’s existing “put” to the shadows of the financial system. This can rust the financial plumbing between banks and non-banks (see Figure 4.5). At least prior to QE, non-banks such as MMMFs had to work hard to get a positive return (ie, higher than bank deposits) by choosing a good counterparty. It is unlikely that MMMFs assets will shrink given the guarantee return from reverse repos (and at odds with proposed regulations – such as floating net asset values (NAVs) – that try to limit the size of MMMFs).

Central banks that have been taking good collateral out of the market for sound macro reasons will not let the ownership of these securities go back to the private market, since it will have an impact on the repo rates (via collateral velocity). However, the market needs the collateral services that these securities can offer, which transfer with possession, not ownership (eg, under the proposed reverse repo, non-banks will get ownership but not possession to reuse securities). Securities in the market’s possession have velocity; those at the central bank do not. There will be a net reduction in overall financial lubrication if non-banks are the primary conduits for the Fed’s reverse repo.

Conclusion

Just as water finds its own level, collateral in the market domain generally finds its economic rent when it is pledged for reuse. The years since Lehman have seen major central banks take out good collateral from markets and replace it with freshly printed money. Sooner or later, these balance sheets will unwind – either voluntarily, when central banks will release collateral and take in money, or involuntarily, as the securities held at central banks mature or roll off. Analytically, the rate of absorbing money will move the LM curve left. Simultaneously, the rate of release of collateral (in lieu of money) will move the IS curve up. So unwinding will increase both the (money) interest rate and the (collateral) repo rate. As both rates move up, policymakers will attempt to keep them close (and not create a wedge between them). This may be another reason why the Fed may want to keep an eye on the repo rate, and hence its reverse repo programme that puts a lid on collateral velocity.

Panel 4.1:How QE Can Jam the Financial Plumbing

Financial Times column, Manmohan Singh, Dec 3, 2014

Central bank asset purchases absorb ‘good’ collateral like Treasury bonds

A layman may not differentiate between central bank actions like quantitative easing, as undertaken by the Federal Reserve, Bank of England and Bank of Japan, and the long-term refinancing operations, or LTROs, used so far by the European Central Bank. However, from a financial plumbing angle the details really matter.

QE absorbs “good” collateral like US Treasuries or UK gilts as central banks buy up the assets. The LTRO, in contrast, is a financing tool for banks where the ECB takes on the not-so-good collateral of the eurozone – for instance bonds that may be trading below face value – while keeping good collateral in the market domain.

Such monetary policy also interfaces with financial plumbing as it straddles the so-called “repo market” where financial institutions use collateral to secure short-term loans. The uncertainty on the Fed’s QE unwind timetable and the release of good collateral may result in repo rates not aligning with the US central bank’s policy rates. On the other hand, the ECB’s two three-year LTRO transactions by definition unwind by February 2015.

The Fed’s balance sheet was still growing until October. But the Fed policy rate is no longer representative of broader financial conditions as it is primarily the negotiated rate between banks that have access to the Fed’s deposit facility paying 25 basis points and non-banks that do not. Market rates have also been kept above zero to support money market funds that must maintain a stable net asset value.

In contrast, the ECB’s balance sheet is shrinking. Eonia, the eurozone’s equivalent to the Fed Funds rate in the US, still responds to factors such as excess liquidity in the eurozone and moves in tandem with target rates that are negative at present. The ECB does not see zero as a hard constraint for either interest rates or repo rates.

The ECB faces other challenges, however. Some of the best eurozone collateral is at the Swiss National Bank, as its balance sheet expanded significantly after the Swiss franc was pegged to the euro in September 2011. This has reduced good collateral circulation in the eurozone, providing lessons for the ECB as it considers a possible QE. More generally, the ECB, as does the Fed, has to consider financial plumbing.

Overall financial lubrication has to be cognisant of both the money and collateral markets, as any collateral shortages that translate into lower (and negative) repo rates will affect the timing and level of the monetary policy rate cycle. So what are some of the lessons for the ECB?

From a financial lubrication angle, markets need both good collateral and money for smooth market functioning and, ultimately, financial stability. Having a ready supply of good collateral like US Treasuries or German Bunds also helps in reallocating the not-so-good collateral.

QE that isolates good collateral from the wider market reduces financial lubrication. Its substitute, money that shows up as excess reserves, is basically contained in a closed circuit system built to avoid inflation by introducing “interest on excess reserves”. Again, in contrast, the ECB is now charging financial institutions 20 bps on excess reserves to encourage money to move to the real economy.

If the ECB embarks on QE, it may help stimulate activity and lift inflation but would also involve risks. For one, it entails an unwind risk that was absent from its LTROs. More importantly, with a larger balance sheet the ECB will need to juggle financial plumbing aspects to avoid creating “wedges”, or differences, between the money rates it controls and collateral rates dictated by the market.

One way to do so is to allow the good collateral associated with QE that comes to the ECB to be simultaneously lent out through an aggressive “securities lending” programme: the securities will still belong to the ECB but will be lent to the market for short periods. That would potentially lessen QE’s detrimental effect on financial lubrication in the wider markets.

Piecing together the lessons from QE, and being aware that the initial conditions for the ECB’s asset purchases are different from those in the US when the Fed embarked on its own programme, it should be obvious that targeted funding to the broader economy in the eurozone is necessary, but may not be sufficient. The banking system will also need to spread central bank monies beyond financial assets to enhance the benefits of a QE programme.

Panel 4.2:Fed’s QE Exit Must Avoid Collateral Damage

Financial Times column, Manmohan Singh, May 22, 2013

With a QE exit now being eyed, many are beginning to worry about the possibility that the Federal Reserve will struggle to unwind its huge balance sheet in a controlled way. In reality, however, there is no reason to fear runaway rates provided authorities keep collateral market rates in mind during the exit process.

Among the unconventional tools the Fed introduced during the peak of the crisis was a “floor” mechanism known as “interest on excess reserves”. Less high-profile than other unconventional policies such as asset purchases, IOER’s role in controlling rates has been underappreciated by the market. Indeed, it is because of IOER that liquidity has been reabsorbed into the central bank system so efficiently. Above all, IOER has played an important role in ensuring that short-term rates did not fall too far below the critical floor of 0.25 basis points, despite all the additional liquidity that was pumped into the system.

Contrary to popular understanding, the Fed has in this sense been propping up the short-term rate market, not suppressing it. lOER’s power to steer rates could now be used to help manage money-market rates during the exit process. But lOER’s ability to influence money market rates for depository institutions has come at a cost. The short-term rates market is increasingly bifurcated, as it is only depository institutions that can really benefit from these freely distributed positive rates. In contrast, non-depository institutions keen on keeping money invested in liquid and safe investments have to fight over an ever-smaller pool of good collateral. This is partly because quantitative easing has sucked a significant sum of quality collateral out of the public market, and partly because the risk-averse are less inclined to part with the quality collateral that remains. Both these changes lead to collateral scarcity in the private money markets, which is important because only the collateral markets can provide deposit-like safe investments, without unsecured bank credit risk, for non-depository institutions. This has caused a disruption in the market for repurchase (repo) agreements.

In a repo trade, an institution either pledges collateral for funding and pays for the privilege to receive those funds, or alternatively – on the other side of the trade – receives collateral as a guarantee for funds lent out. Collateral shortage lowers the repo rate; collateral abundance increases it. In today’s market, repo rates arguably represent the true cost of money for non-depository institutions that do not have access to the Fed’s reserve system. A central bank such as the Fed can try to dictate the cost of money by setting target interest rates for unsecured funds, but if the repo markets do not comply, the central bank can in some sense be judged to have lost control.

While the Fed has managed to influence the cost of Fed funds for depository institutions – in large part due to IOER it has found it much harder to ensure those rates are available to all parts of the market. In fact large cash positions of such non-banks in the midst of collateral shortage have now started to influence the key Fed funds rate, seeing it trade below the IOER “floor”.

But it is not just the Fed that has been suffering from such collateral pressures. In the eurozone, collateral scarcity in German, French and Dutch short tenor bonds has even seen rates turn negative. The same has also been witnessed in Danish and Swiss bonds. So far the US has managed to avoid rates turning negative because IOER has helped to maintain short-end rates – both money and repo. It has been successful precisely because the average repo rate has tended to be lower than the IOER rate. It is true that if the Fed was to release collateral and take in money as part of its unwinding process, there is a risk that the repo rate could begin to increase and start to influence market rates widely. However, provided the average repo rate remains near IOER, there is no real reason why other rates should rise precipitously, given the IOER incentive for depository institutions to keep liquidity parked at the central bank.

The repo rate overshooting IOER can lead to inflation, or expectations thereof, since depository institutions may switch liquidity from IOER to repo markets, thereby increasing the money multiplier. This is why any unwinding should be focused on steering collateral to those non-bank areas that are currently suffering shortages, depressing rates, rather than to depository institutions whose cost of money can continue to be much more effectively controlled by using IOER. The Fed has been developing tools to do just that, including a programme of trading repo directly with large money market funds. (But then came the reverse repo(s) in September 2013, discussed in this chapter.)

Panel 4.3:The Fed, Reverse Repos and Tri-Party/Bilateral Repo Market Wedges

This is a guest post by Manmohan Singh, a senior economist at the IMF, in FT Alphaville, the FT’s digital news and commentary service, February 14, 2014. Views expressed are his own and not those of the IMF.

The idea of eliminating the present wedge between Fed’s reverse repo program (RRP) floor and the interest on excess reserves (IOER) is intriguing because such a change would only allow the Fed to set the price on such operation (P), and would leave the market to determine the quantity of reserves (Q) on Fed’s balance sheet (Gagnon/Sack proposal).

As background, Quantitative Easing has greatly increased banks’ holding of reserve balances at the Federal Reserve. This is primarily due to nonbank deposits with banks, since most of QE was between Fed and nonbanks (re: Carpenter et al, 2013). In other words, QE converted good collateral to excess balance sheet at banks. To the extent that banks face leverage ratio constraints as a result of QE, they want balance sheet “space” for financial intermediation/non-depository activities. At the same time, regulatory changes are boosting demand for high quality liquid assets (HQLA). The discussion below focuses on financial plumbing and possible wedges between rates in the Triparty repo and bilateral repo markets.

The operational structure of the RRP facility puts practical restrictions on the reuse of collateral outside the Triparty system. Thus focusing on the liability side, for every $x billion of reverse repos, the line item RRP on the liability side of Fed’s balance sheet will go up by $x billion dollars, and bank excess reserves, also on liability side, will go down by $x billion dollars. The Fed’s total balance sheet is unchanged, but bank balance sheets in total will likely shrink as reserves decline. Banks (and members of government securities division of DTCC) can reuse the collateral within the Triparty system. The “released” collateral stays on the Fed’s balance sheet or within the Triparty system (hence, “capped rehypothecation”). In other words even if bids for RRP were uncapped, collateral will be contained and not freely available to the financial system. Within the present Triparty structure, none of the collateral can be used to post at central clearinghouses, or in the bilateral derivatives markets. In general, securities in the market’s possession are reused and have velocity; those at the central bank do not.

The bilateral repo market in the US that is the core of the bank/non-bank nexus is outside the Triparty framework. In a recent speech, Fed Governor Tarullo mentioned the size of the bilateral repo market at $1 trillion, which presumably would be higher if the velocity of collateral is factored in. The bank balance sheet space opened up by nonbank RRPs with the Fed could make this possible, enhancing the link between the Triparty and bilateral repo markets. Demand from the bilateral repo market may entice some banks – if they have balance sheet space (after adjusting for HQLA/leverage ratio/LCR) – to make a market for certain clients like pension funds/insurers that are not eligible for access to RRP but would like to obtain high quality collateral. This demand may lead to banks undertaking collateral transformation (including substitution of their balance sheet collateral with RRP collateral) and is at the core of financial intermediation. Without the Tri-party features associated with Fed’s RRP and given the size of the bilateral repo market, collateral velocity could increase, leading to a wedge between bilateral repo rates and the RRP rate.

As an example, a bank that has surplus money can lend to the Fed, collateralized under the RRP (assume at 25 bps). Or, lend to a hedge fund at 30 bps, collateralized. Here the bank/hedge fund bilateral repo rate is above the 25 bps RRP. Alternately, this bank may have surplus HQLA (or get it via RRP) earning 25 bps. This collateral may be in demand by a pension fund to post at a CCP. The repo rate between pension fund/bank will not be more than 20 bps (perhaps even 10 bps, factoring the FDIC levy to the bank), as the bank takes its “cut”. This wedge around the 25 bps RRP (30 bps to 10 bps) can only be removed if hedge funds can deal with the pension fund directly; but they don’t and this is where financial intermediation comes in.

Also, the term deposit facility (TDF) allows banks to shift reserve balances to term deposits with the Fed, and might lead to a shift of holdings from large banks to small banks, thereby creating space for the former. The TDF should not have a significant impact on collateral dynamics, and the smaller banks may have the balance sheet capacity to absorb liquidity at rates modestly above 25 bps paid on reserves.

ECB did not resort to a floor when repo rates turned negative. Since ECB cut deposit rate to zero in July 2012, for much of the time, repo rates of good collateral (German Bunds, French Oats) remained below zero. EONIA (the key money market rate) is now in positive territory as excess liquidity declines with LTRO repayments; so good collateral repo rates also move positive. Neither does UK provide a floor to repo rates; their bank rate, similar to Fed’s IOER, is at 50 bps. [Last September, the Fed started a trial program “testing” a repo rate floor via the RRP. In fact even with 25 bps rate (IOER) for banks that pulls repo rates up, Fed “put” a 5 bps floor. This leads to an asymmetry in distribution to the savers in the real economy and benefits short-term investors].

In summary, the Fed’s exit strategy suggests tapering of purchases and ultimately their cessation, interest rate lift-off (with clarification of what the policy rate will be), and maybe also asset sales down the road. This exit strategy needs to be mindful of disruptions to the financial plumbing due to the possibility that a sizable (and quick) reduction in reserve balances could lead to wedges between the bilateral repo rate and the rate on Fed’s RRP operations. Specifically, as balance sheet space is created via RRPs, the demand for safe assets from entities outside the Triparty system and the potential for non-depository activities of banks should not be underestimated.

References

Despite the European Central Bank’s (ECB’s) efforts to take in lower-grade collateral, actions of the Swiss National Bank (SNB) (and other central banks) have unintended consequences and diluted this ECB objective. Largely due to the Swiss franc/euro peg, the SNB balance sheet is now over €500 billion with about half of the assets comprising “core” euro bonds and equities. Despite the growth in the SNB balance sheet, the asset allocation as a percentage towards core euro bonds has remained unchanged (although a bit lower now – at 42% – than before). Although in line with past asset allocation trends, SNB’s bond purchases withdraw the best and most liquid collateral from the eurozone; this reduces the collateral reuse rate, since these bonds are siloed at the SNB and not pledged in the financial markets. Siloed collateral has zero velocity by definition.

In the US the supplementary leverage ratio is higher than the leverage ratio in Europe but not binding until January 1, 2018. So in the interim, the balance sheet space constraint is rather a soft-constraint. This softness allows banks to take non-banks’ money, and split the 25bp IOER with them now the IOER is 50bp, after the Fed’s lift-off on December 1, 2015. This, among other reasons, keeps the volume in the Fed Funds market from collapsing.

    Other Resources Citing This Publication