Blog

With this week’s further cut, the Federal Reserve has brought its target Fed funds rate below the annual increase in the “core” personal consumption price index (2.0% versus 2.1%). Historically, this measure of real interest rates has become negative only well into or after recessions – see first chart. Indeed, in two cases – the 1980 and 1981-82 recessions – real rates troughed above zero.

With the advance first-quarter GDP estimate suggesting marginal growth (0.6% annualised), a recession has yet to be confirmed – see also here. Intrade’s recession contract now implies a 33% probability of two consecutive negative GDP quarters in 2008, down from a recent peak of 79%.

The Fed normally requires evidence of significant economic slack before easing policy aggressively. Both the unemployment rate and industrial capacity utilisation are currently little different from their averages over the last five years.

A useful survey-based measure of industrial capacity pressures is the ISM manufacturing “supplier deliveries” index, based on the number of purchasing managers reporting longer delays in obtaining production inputs. An index level of 52 has historically separated periods of Fed easing and tightening – see second chart.

The index was modestly below 52 between February and November last year, signalling rate cuts, although not to the extent actually delivered. It has recovered more recently, reaching a 20-month high of 54 in the April survey released yesterday.

The current divergence between capacity use indicators and the level of real interest rates is unsustainable. Unless capacity pressures ease significantly, the Fed may be forced to consider reversing recent cuts later in 2008.

Inflationary pressures often strengthen in the early stages of economic slowdowns, for at least four reasons. First, output is typically above its trend or potential level when the slowdown begins; a fall beneath trend is needed to stem inflation momentum. Secondly, productivity tends to slow along with output, as employers are initially reluctant to cut workforces, implying faster growth in unit labour costs. Thirdly, monetary expansion often remains strong well into an economic downswing; ample liquidity accommodates price increases and may boost inflation expectations. Fourthly, a slowing economy may be associated with a fall in the exchange rate, putting upward pressure on import prices. This last effect can be particularly important if world output is still growing solidly, pushing up prices of tradable goods and raw materials.

Forecasters who expect the current inflation overshoot to prove short-lived, implying scope for the Monetary Policy Committee to continue to reduce Bank rate, emphasise the likely moderating impact of a widening negative “output gap”. Estimates of the gap are never uncontroversial but even inflation pessimists concede that output is unlikely to be far above its potential level at present. OECD calculations are probably representative of the consensus, suggesting a positive deviation of just 0.1% in the first quarter. Forecasters expect GDP to grow by just 1.1% in the year to the fourth quarter of 2008 and by 2.1% in the subsequent year, according to Consensus Economics. Based on the OECD’s estimate of potential growth of 2.7% pa in 2008 and 2009, this implies a negative output gap of 1.3% by this year’s fourth quarter and 2.0% by late 2009.

The impact on inflation of rising slack may, however, be outweighed by the three other factors cited above. Import prices soared 10.4% in the year to February, the largest annual gain since 1993, following sterling’s expulsion from the ERM. With the effective rate down 3% since February and global commodity prices rising further, the pick-up is likely to continue. Monetary expansion also remains robust, despite tighter credit conditions: M4 rose 11.9% in the year to March and has outpaced nominal GDP by a cumulative 22% over the last three years – slightly greater than the peak differential reached in the late 1980s Lawson boom. Meanwhile, productivity is slowing as the economy cools: aggregate hours worked rose an annual 1.0% in the three months to February, up from 0.4% a year earlier, despite a fall in GDP growth over the same period.

To investigate these issues further two models were estimated, the first explaining inflation by its own history, GDP growth, the output gap and indirect tax changes, the second also incorporating import prices and M4 expansion. The price measure used was the domestic expenditure deflator – covering investment and government spending as well as private consumption – and the models were estimated on data back to the early 1970s, when sterling was floated. Forecasts were generated based on assumptions about the inputs. Specifically, GDP growth and the output gap were assumed to follow the path described above, import prices to stabilise from the second quarter of 2008 and M4 expansion to moderate steadily, reaching an annual 6% by late 2009.

The results are shown in the chart. The simpler model excluding import prices and M4 suggests annual inflation will fluctuate around 3% during 2008 before embarking on a sustained decline in 2009 and beyond, reflecting the restraining impact of a widening negative output gap. By contrast, the extended model forecasts a further sharp rise in inflation over the coming year, with recent import price gains and excessive M4 growth offsetting the output gap effect until late 2009. The extent of the near-term increase is probably exaggerated, since the estimation period includes the 1970s and 1980s, when monetary policy had less credibility and inflation expectations were therefore more volatile. Even allowing for some overstatement, however, the forecast suggests inflation will rise by further and for longer than most economists expect.

Component-level analysis of the consumer price index confirms the possibility of an extended inflation overshoot over coming months. Energy suppliers have raised electricity and gas tariffs by 14% since the start of 2008, reflecting wholesale price rises during 2007. One-month sterling forward prices of Brent crude and natural gas have climbed 33% and 24% respectively since the fourth quarter. A further rise in household tariffs looks inevitable if these levels are sustained; a 10% gain would ensure annual CPI inflation moves well above the 3% upper target limit later in 2008, sustaining the breach for several months. Additional risks stem from pass-through of recent food and import price increases and a possible effort by retailers to rebuild margins.

The above analysis questions consensus hopes of a steady progression lower in Bank rate over the remainder of 2008. Further cuts are possible but are likely to require dramatically weaker economic news. The MPC will also need to calibrate any action against money and credit market developments; a continuation of the recent small improvement – particularly a fall in LIBOR / Bank rate spreads – would substitute for policy easing. Indeed, a rapid normalisation of market conditions could leave the current level of Bank rate looking too low relative to prospective inflation.

The chart below shows an updated profile for annual consumer price inflation over the next 12 months based on the following assumptions:

1. Household electricity and gas tariffs are raised by a further 10% over the summer in response to recent steep rises in wholesale energy costs. This would imply a full-year increase of 25%.
2. Unprocessed food price inflation is stable at its current annual rate of 2.4%.
3. “Core” consumer prices – i.e. excluding unprocessed food and energy – rise at a 2.25% annual pace for the remainder of 2008, falling to 2.0% during 2009. This is modestly higher than over the last 12 months, reflecting assumed pass-through of large import and raw material cost increases.

On these assumptions, which do not appear unduly aggressive, headline CPI inflation is projected to be above 3% in every month between July 2008 and January 2009, peaking at 3.5% in September.

If correct, such a profile would raise a large question mark over consensus hopes that the MPC will lower Bank rate steadily over the remainder of 2008.

As discussed in earlier posts, the preferred version of my US recession probability indicator (i.e. incorporating credit spreads) rose sharply in 2007 but peaked just short of the 50% “trigger” level. It suggested the economy would be very weak in late 2007 and early 2008 but would skirt an official recession, as determined well after the event by the National Bureau of Economic Research (NBER).

Amid widespread gloom, it may seem perverse to claim a recession is still not a done deal. However, available evidence suggests next week’s first-quarter GDP report will show modest growth, of perhaps 1% annualised, with gains in net exports and stockbuilding offsetting a further slump in housebuilding and flat consumption and business investment.

Aggregate hours worked by private non-farm employees declined at a 1.2% annualised pace in the first quarter. Allowing for trend productivity growth of 2%-2.5% pa, this also looks consistent with a small rise in output.

The consensus expects GDP to contract in the second quarter but timely coincident indicators have yet to suggest a further loss of momentum. New jobless claims have stabilised since late March, with only one recent week exceeding the 400,000 level suggestive of recession. As noted by Trim Tabs, withheld employment tax receipts also imply labour market resilience – see first chart below.

Recent further energy price gains will squeeze budgets but households are due to receive tax relief of over $100 billion – 1% of annual disposable income – starting next month. Investment will be supported by $50 billion of incentives for firms to purchase equipment in 2008.

Markets appear to be a little less certain of recession. Intrade’s contract allowing bets on the probability of two consecutive negative GDP quarters in 2008 has recently fallen back from a peak of 79%. (Two negative quarters is arguably a tougher requirement than an official recession – the NBER bases its determination on a range of indicators, not just quarterly GDP.)

For investors, whether the economy is in recession currently is less important than its position in six months’ time. The second chart updates the recession probability indicator, which is giving an “all-clear” signal for late 2008, based on recent falls in real interest rates and a steeper yield curve, which are judged to outweigh tighter credit conditions. (I have made some minor changes to the indicator to improve its historical performance. The new version would have predicted the eight recessions since 1955 with no false signals. Its recent peak remains just below the 50% “trigger” level.)

Simple models may well prove fallible in current unusual times but I continue to expect the US economy to perform better than many fear in 2008, with greater risk of disappointment in Europe. 2009 may be another story, however.

The fees payable under the Bank of England’s special liquidity scheme (SLS) imply the facility will be attractive only to banks currently either unable to borrow longer-term funds at LIBOR or unable to do so in sufficient size. The reported high take-up under the scheme – an initial £50 billion, expected to rise significantly further – is therefore surprising.

One explanation, as mentioned earlier, is that larger, unstressed banks have agreed to participate so that weaker institutions are not stigmatised for using the scheme. Banks may also be hoping that a high take-up will boost perceptions that the scheme represents a solution to the liquidity crisis and so help to restore confidence in the banking system, contributing to a fall in risk premia.

An alternative, more worrying possibility is that the problem of banks being unable to access the interbank market on reasonable terms is more widespread than previously thought. This would support claims that LIBOR seriously understates the true cost of funds for most institutions.

In this latter case, the SLS could be of significant benefit in reducing funding costs down to the quoted level of LIBOR. However, it remains unclear why it should result in a significant fall in the LIBOR-Bank rate spread.

For mortgage borrowers seeking to refinance fixed-rate loans, any helpful impact of the SLS may be outweighed by a recent sharp rise in short gilt yields and interbank swap rates, mainly due to international factors. As the chart shows, the two-year swap rate – a key influence on short-term fixed-rate mortgage pricing – has climbed 40 basis points since the April MPC meeting, reaching a four-month high.

My MPC-ometer suggested a 6-3 vote for a quarter-point rate cut in April, in line with the Sunday Times Shadow MPC. Today’s minutes reveal a 6-2-1 split, with Tim Besley and Andrew Sentance voting for no change and David Blanchflower for a half-point move.

There has been only one previous occasion of a three-way split in a month when the MPC has changed interest rates – January 1999, when rates were again cut by a quarter-point. Three-way splits also occurred in May and August 1998 and May 2006; rates were left unchanged in all three cases.

The minutes imply a further cut could be delayed until July or even later. Besley and Sentance will remain reluctant to ease aggressively until inflation expectations moderate – unlikely before late 2008. Within the majority group, some members voted for a reduction this month to keep policy consistent with the gradual easing path implied by the February Inflation Report projections. On this view, additional action may not be needed until August.

There are also hints of greater concern about exchange rate weakness, arguing for going slow on rate cuts. According to the minutes, a fall in the real exchange rate caused by an increase in the risk premium on sterling assets “would tend to boost net trade and warrant a more pronounced slowdown in domestic demand growth in the near term than otherwise”.

With inflation risks rising, the MPC is right to discourage hopes of rapid rate cuts but sharply weaker economic news or further financial woes could yet force the Committee’s hand.

The Bank of England’s new “special liquidity scheme” may prove ineffective because of its unattractive fee structure.

The scheme appears to be in the spirit of the term auctions of three-month funds last autumn . These attracted no bidders because the Bank set the minimum bid rate at a penal level.

The fee payable on a swap of mortgage-backed securities for Treasury bills will be the spread between three-month LIBOR and the three-month gilt repo rate – currently about 100 basis points. A floor has been set at 20 basis points but is irrelevant. Banks will use Treasury bills to obtain funds in the market at the gilt repo rate. Their total cost of funding – including the fee – will therefore equal LIBOR.

The scheme will help any institutions currently unable to access interbank funds at LIBOR but seems unlikely to result in a significant reduction in the current wide LIBOR-Bank rate spread. There is a danger that use of the scheme will be interpreted as an admission of weakness. The major banks may have agreed to participate in the scheme regardless of their need for funds to mute this signalling effect.

The scheme differs significantly from the Federal Reserve’s term securities lending facility, in which fees are set in an auction subject to a minimum for AAA-rated asset-backed securities of 25 basis points. Details of the latest auction show that some participants paid the minimum fee, implying the swap will have allowed them to obtain funds in the market significantly below LIBOR. The Bank of England appears to have rejected a comparable fee structure because of “moral hazard” concerns.

The relatively restrictive nature of the scheme suggests the LIBOR-Bank rate spread will remain elevated and the onus will be on the Bank’s Monetary Policy Committee to bring market rates down via further cuts in its Bank rate .

—–
COMMENT:
AUTHOR: Julian D. A. Wiseman
EMAIL: jdaw1@the-domain-of-linked-story.com
IP: 68.161.173.136
URL: http://www.jdawiseman.com/papers/finmkts/stigmatisation_t_bills.html
DATE: 04/22/2008 11:40:29 AM

The fee is indeed large. So large that the facility will be useful only to banks unable to borrow enough at Libor. Hence market participants will know that those holding lots of T-Bills, whether selling them or using as collateral, are the weak banks unable to borrow. The linked article discusses this possible stigmatisation of UK T-Bills:

http://www.jdawiseman.com/papers/finmkts/stigmatisation_t_bills.html

The equivalent yield on prime commercial property rose further to 5.9% in the first quarter and is now 110 basis points above its recent trough, according to CB Richard Ellis. The yield remains below its average of 6.4% over 1972-2007 but this may be a poor guide to “fair value”, for two reasons.

First, rents fluctuate significantly with the economic cycle. A high yield may not indicate that property is cheap if rents have been boosted above a sustainable level by a buoyant economy. Conversely, it may be right to invest when yields are low if rents are below trend and likely to benefit from future strong economic growth.

Secondly, any judgement about valuation must take account of returns on competing assets. The rental yield is often compared with yields on conventional gilts but this is invalid because bond interest is fixed while rents rise with inflation over the long run. In other words, the rental yield should be compared with real not nominal interest rates.

The chart shows a measure of valuation that incorporates these considerations – the gap between the normalised or cyclically-adjusted rental yield and real yields on long-term index-linked gilts. The normalised yield is currently lower than the actual yield (5.6% versus 5.9%) because rents are estimated to be 4% above trend, reflecting the economy’s recent strength. (There were much greater deviations in the early 1970s and late 1980s, when rents overshot by 30-40%.)

A below-average normalised yield is, however, counterbalanced by the low level of yields on long-term index-linked gilts. The gap between the two has surged from 3.1% in last year’s second quarter to 4.7% currently – the highest since 2004 and well above a long-term average of 3.6%.

Tight credit conditions and a slowing economy will restrain demand for property space but current valuations already discount much gloom. Property should outperform government bonds over the medium term.

Markets are awaiting details of new measures promised by the Bank of England to ease banks’ difficulties obtaining longer-term funding, particularly to finance mortgage lending. What form could they take?

In principle, the Bank can offer relief in three main ways:

1. Increase the average term of its lending to the banking system.
2. Widen the definition of eligible collateral against which loans are made.
3. Increase the total volume of lending.

Under 1, the Bank has already moved a long way. Longer-term lending (i.e. for three months or longer, including the loan to Northern Rock) now accounts for an estimated 80% of funds advanced to the banking system, up from 30% in September last year. A further increase is possible but is unlikely to have a significant beneficial impact.

Under 2, the Bank has allowed some widening but continues to operate stricter rules than the Fed and ECB. It still requires government collateral for its normal weekly operations and a portion of its longer-term lending. Banks have been allowed to use AAA-rated asset-backed securities in special auctions of three-month funds in December and January, which have been rolled over and expanded in March and April. Last September, the Bank offered to auction three-month funds against a broad range of collateral including mortgages and corporate bonds as well as ABS but imposed a high minimum bid rate, resulting in no take-up. Applying this broad collateral definition to all longer-term lending, without enforcing a penal rate, would be helpful in both widening access to official funds and allowing liquidity-short banks to borrow in greater amounts.

A significant improvement in funding conditions is, however, also likely to require measure 3 – an increase in aggregate lending to the banking system. This presents a technical issue: any such expansion requires offsetting sterilisation measures to prevent banks’ reserve balances with the Bank rising above target. (An overshoot of reserve balances would result in very short-term interest rates falling below Bank rate, undermining MPC policy.) Two possible measures for achieving a rise in lending without boosting banks’ reserves are as follows. First, the Bank could ask the Debt Management Office to repay immediately the government’s remaining “ways and means” borrowing from the Bank, releasing funds for market operations. Secondly, the DMO could issue an additional quantity of Treasury bills or gilts relative to its current plans, placing the proceeds on deposit at the Bank for onward lending to the banks.

The tables below illustrate how the Bank of England’s balance sheet and the size and composition of its lending to the banking system might change if these proposals were implemented. Specifically, the following assumptions are made:

1. A further reduction of £3 billion in short-term lending in favour of longer-term loans (measure 1 above).
2. Application of the broad September definition of eligible collateral (i.e. including mortgages) in all longer-term operations (measure 2).
3. Full repayment of the remaining £7 billion “ways and means” advance to the government.
4. Placement of a £20 billion special deposit by the DMO at the Bank, financed by additional issuance of gilts and / or Treasury bills.

Changes c and d would allow an increase in the Bank’s aggregate lending to the banking system from its current level of £65 billion to £93 billion. Within this total, changes a and b would permit new longer-term lending of £38 billion against mortgage collateral. This is a significant sum in the context of the overall mortgage market – sufficient to finance five months worth of net lending at its recent pace.

In addition to these measures, the authorities should also consider emulating the Federal Reserve’s “term securities lending facility”, under which banks are able to swap mortgage-backed securities for Treasuries held on the Fed’s balance sheet, with the Treasuries then used as collateral to obtain funds in the market. The Bank of England holds few gilts on its balance sheet so such a facility would require the DMO to create extra gilts specifically for this purpose.

—–
COMMENT:
AUTHOR: Interested
EMAIL:
IP: 77.98.197.155
URL:
DATE: 04/17/2008 09:42:44 AM

Simon
I wonder if you would comment on the fact that the BOE now seems to being doing for all UK banks what it said it could not do for the then management of Northern Rock two months ago. Namely opening up a brand new line of funding. As I understand it the banks still have access to the lender of last resort at the BOE but they are unwilling to use it as they have seen what happened to Northern Rock.

Could you also comment on the fact that given much of the problem with Northern Rock was caused by the bank being named and it appears it was simply the first UK bank to need the funding and the rest all need it now. If you think it is right that Northern Rock followed the rules and was nationalised yet obviously many banks are now getting extra BOE funding and are not being named? Do you think this will have any affect on possible compensation paid to the then shareholders of Northern Rock?

Thank you

Interested

—–
COMMENT:
AUTHOR: Simon Ward
DATE: 04/18/2008 09:56:41 AM

There is a chance that Northern Rock could have survived as an independent institution if it had enjoyed access to ECB-type liquidity support or the new Bank of England swap facility currently under discussion. However, Rock’s management should have made plans on the basis of existing UK arrangements. Its business model is widely acknowledged to have been extreme. Access to ECB funds did not prevent the failure of the German banks IKB and Sachsen last year.

Press reports suggest that the planned facility to allow banks to swap mortgage assets for gilts could be as large as £100 billion. If true, this would make it much more significant that the Federal Reserve’s equivalent scheme – the term securities lending facility (TSLF).

The TSLF is currently capped at $200 billion, of which $100 billion had been used as of Wednesday 9 April. $200 billion is equivalent to 1.8% of the US broad money supply (expanded M2). The same percentage of UK broad money M4 would be £29 billion.

The UK scheme arguably needs to be larger because the Fed has introduced a range of other measures to help markets. Total Fed support amounts to $470 billion* or 4.1% of broad money. The equivalent UK sum would be £69 billion. From this should be subtracted the £25 billion of three-month funding the Bank of England has offered to banks against AAA-rated asset-backed securities. This suggests the UK swap facility should be £44 billion to equate total UK and US support. (An earlier post suggested additional aid of about £40 billion was necessary to ease mortgage market strains.)

Demands for a £100 billion facility seem excessive. The success of the scheme will depend as much on the precise collateral rules and term of the swap as its size.

* Breakdown: TSLF $200 billion, term auction facility $100 billion, additional term repos $100 billion, swap facilities with ECB and Swiss National Bank $36 billion, primary credit discount window lending $7 billion (9 April), primary dealer credit facility lending $26 billion (9 April).