Broad money figures for March show a disappointing initial impact from QE. It is too early to make a firm judgement but the numbers suggest that Bank of England asset purchases will need to be expanded beyond the planned £75 billion by early June to boost monetary growth sufficiently to support a sustained economic recovery. (The Bank has Treasury authority to buy up to £150 billion.)

Monetary trends are best monitored using the Bank of England’s adjusted M4 measure excluding “intermediate other financial corporations” – this removes distortions due to the financial crisis but includes investing institutions’ money holdings, which should be inflated by a successful QE operation. The Bank of England does not release its monthly adjusted M4 estimates but the aggregate is likely to have been little changed in March, since overall M4 rose by only 0.2% on the month, while M4 excluding all financial corporations declined by 0.1%. (The Bank will publish March quarterly data for its adjusted M4 and M4 lending measures on 8 May.)

The Bank of England bought £15 billion of gilts in March, equivalent to 0.9% of adjusted M4. There are two possible reasons why the money numbers have shown little response. First, the Bank may have bought securities mainly from banks and overseas investors rather than domestic non-bank investors – only money holdings of the latter are included in M4. Secondly, a positive impact from QE may have been offset by other influences on M4, such as weak bank lending to the private sector.

Both factors were in play in March. Statistics on gilt transactions by sector show that Bank purchases of £15 billion were balanced by net sales of £7 billion by overseas investors, £6 billion by domestic non-banks and £2 billion by banks – see table. So the direct impact of purchases on M4 was only 40% (i.e. £6 billion out of £15 billion). (The outstanding stock of gilts was little changed in March, with DMO issuance offset by a large redemption. This may have affected the sectoral pattern of transactions.)

Meanwhile, credit weakness was a significant drag on M4: sterling bank lending to households and non-financial corporations, adjusted for securitisations, rose by only 0.2% in March and there was a large repayment of net foreign currency borrowing by UK residents. In addition, while domestic non-bank investors reduced their gilt holdings by £6 billion, they bought £21 billion of other forms of public sector debt, including £10 billion of Treasury bills. The overall public sector contribution to monetary growth – including the impact of the public sector net cash requirement, the Bank’s purchases and other debt transactions – was therefore only £6 billion, or 0.4% of adjusted M4.

While this month’s numbers are disappointing, money trends have improved significantly since last autumn: M4 excluding financial corporations rose at a 4.0% annualised pace in the three months to March after just 0.8% during the the fourth quarter. Annual growth slipped to 2.5% in March but in real terms – relative to retail price inflation – has recovered from a low of -0.6% in October to 2.9% now. This is consistent with improving economic prospects but a further pick-up is needed to lay the monetary foundations for a recovery.

The MPC is likely to wait until its June meeting before deciding to expand its QE operation, partly because it would be unwise to make a judgement about its impact on the basis of just one month’s data, and partly because an expansion of the programme would probably take the form of an extension of buying for a further three months, rather than a step-up in the weekly pace of purchases.

Eurozone monetary statistics for March and the latest survey of bank loan officers suggest an urgent need for the ECB to embrace US / UK-style quantitative easing at its meeting next week.

Broad money M3 has contracted over the last three months, pulling annual growth down to a five-year low of 5.1% – see first chart. The liquidity squeeze remains focused on the corporate sector, with M3 deposits of non-financial companies 1.2% lower than a year ago.

In terms of the credit counterparts, M3 weakness reflects a recent fall in bank lending to the private sector – second chart. A similar decline in the US has been offset by the expansionary impact of the Fed’s securities purchases, so US M2 has continued to grow, albeit at a slower pace than in late 2008 – see last post.


The latest bank loan officer survey shows a fall in the net percentage tightening credit standards on corporate loans but the decline was much less than in the Bank of England survey released in early April – third chart. The equivalent US survey is due next week; the Fed’s statement yesterday referring to “some easing of financial conditions” hints at favourable results.

US money measures accelerated sharply when the Federal Reserve embarked on quantitative easing in late 2008, buying commercial paper and mortgage-backed securities. Three-month growth in the broader M2 aggregate reached an annualised 24% in December – see first chart. The monetary injection laid the foundations for the March / April rally in equities and recent improved economic news.

M2, however, began slowing in early 2009 and has actually fallen over the last four weeks, bringing the three-month rate of change down to 3%. Annual growth remains solid at 8% but has retreated from 10% in January.

Recent M2 weakness has not reflected any slowdown in Fed securities purchases. At its March meeting, the Fed announced a big expansion of its buying plans to a potential $2 trillion by the end of 2009 – second chart. This would imply $1.25 trillion of purchases over the remaining eight months of the year, or about $35 billion per week. Recent buying has been on roughly this scale.

Unlike the ECB and Bank of England, the Fed does not publish a “counterparts analysis” of the drivers of M2 growth but it appears that the expansionary impact of official securities purchases has been offset by a recent contraction in bank lending to the private sector. Commercial bank loans and leases outstanding have fallen at an annualised 5% rate over the last three months – third chart.

On further analysis, this contraction mainly reflects declines in commercial and industrial loans and advances under sale-and-repurchase agreements. Corporate lending has been depressed by recent heavy destocking while the fall in repo advances is consistent with other evidence of investor deleveraging. With the stocks cycle turning, and investor risk appetite beginning to revive, lending trends could improve going forward.

M2 trends are not yet ringing alarm bells but a further slowdown would question the sustainability of recent equity market gains and tentative economic improvement.

The Treasury’s medium-term fiscal forecasts appear to rest on optimistic assumptions about future borrowing costs. On reasonable alternative assumptions, public sector net interest payments could rise to 3.9% of GDP by 2013-14 versus an official projection of 3.0%.

The Treasury’s forecasts for debt interest have received limited scrutiny partly because they are buried within the detail of the Budget documents. Medium-term projections for public sector net interest as a percentage of GDP can be derived from Table C2 of the Financial Statement and Budget Report (FSBR) as the difference between public sector net borrowing and the “primary balance”. These forecasts can be converted into nominal terms using money GDP assumptions from Table C1.

To derive the Treasury’s unpublished assumptions about future borrowing costs, it is necessary to put these public sector net interest numbers onto a general government gross basis by adding back estimated interest receipts and adjusting for public corporations. The gross interest projections can then be compared with published numbers on gross government debt to derive an average interest yield.

According to the FSBR, public sector net interest will rise from 1.6% of GDP in 2009-10 to 2.6% in 2010-11 and 3.0% in 2011-12 – see first chart. A further small increase to 3.1% in 2012-13 is then reversed in 2013-14, when the proportion returns to 3.0%. This stabilisation raises suspicion, since general government gross debt is projected to rise by 17% in the two years to the end of 2013-14.

To generate this profile, the Treasury must be assuming a fall in the interest yield on government debt in 2012-13 and 2013-14. The FSBR projections are consistent with the yield averaging less than 4% in the three years to 2013-14, far below projected money GDP growth of more than 6% per annum over this period – second chart.

The bulk of outstanding debt consists of fixed-coupon gilts. The interest yield in a particular year is a weighted average of the rates on existing debt and new borrowing – not just to cover the budget deficit but also to finance gilt redemptions and roll over the stock of Treasury bills. For the average yield to fall after 2011-12, as implied by the FSBR forecasts, the interest rate on new borrowing in 2012-13 and 2013-14 would have to be well below 4% – a rough calculation suggests an average of about 3% over the two years.

The charts present an alternative scenario for the average yield and net interest as a percentage of GDP based on the assumption that the interest rate paid on new borrowing in 2011-12, 2012-13 and 2013-14 is equal to the projected rate of money GDP growth in each year (i.e. 6.0%, 6.2% and 6.1% respectively). This generates a gradual rise in the interest yield to 4.9% by 2013-14. While significantly higher than the 3.8% implied by the Treasury’s forecast, this is below the 5.2% average between 2004-05 and 2007-08 (when money GDP grew more slowly than projected for the three years to 2013-14).

On this alternative scenario for the average yield, net interest as percentage of GDP rises to 3.9% of GDP by 2013-14 against the Treasury’s projection of 3.0%. More pessimistic scenarios are clearly feasible, based on investor concerns about inflation and / or solvency pushing new borrowing costs above the rate of money GDP growth.

A higher interest bill would imply either a greater squeeze on non-interest spending or, more likely, further fiscal slippage. According to the Treasury’s forecasts, real current spending will rise by 0.7% per annum in the three years to 2013-14. Stripping out interest costs, however, the rate of growth is just 0.2% pa. On the alternative scenario presented here, real non-interest spending would need to fall by 0.7% pa over this period to make room for higher debt-servicing costs, assuming no further upward revision to plans.