During his tenure at the helm of the UK Debt Management Office (DMO), Robert Stheeman has certainly lived through interesting times. If this has been a curse, his cheerful demeanour and the matter-of-fact way he recalls tumultuous events do not suggest it. When he took the post in 2003 the UK raised £26.3bn ($41.8bn) of debt in 13 Gilt auctions. In 2009-2010 the DMO tapped the market in 58 auctions, 12 mini-tenders, six syndications and one other tender. Some £227.6bn ($361.8bn) of revenue flowed to the coffers of Her Majestys Treasury.
Says Stheeman: "The speed at which the numbers changed was the truly extraordinary thing. At the end of our 2007-2008 year [March] we issued £58.5bn. Thirteen months later our target was £227bn. I did not predict the financial crisis. But even if I had, if someone had said that would be the amount of Gilts the DMO would issue in a single year when I joined in 2003, I probably would have told them to sit down, reflect calmly and let the moment pass."
Stheeman praises his colleagues at the DMO. But he also acknowledges these are far from normal times. By the time that the Bank of England finishes its latest round of Quantitative Easing (QE) in November, it will have purchased £375bn of Gilts from the secondary market. One buyer owns approximately a third of the entire market of nominal Gilts.
Quantitative easing has done what it was intended to do, drive Gilt yields down.
But as a consequence few investors now think Gilts represent good value. "Gilts are a seriously unattractive investment at these levels. From a personal perspective there is no way I would hold Gilts," says Marino Valensise, chief investment officer at Baring Asset Management, who oversees $60bn. "There is little space for capital appreciation and a lot of potential downside."
Andrew Wickham, head of UK fixed income, at Insight Investment, a specialist in liability-driven investment and fixed income with £182bn ($291bn) of assets under management, echoes this sentiment. "It is hard to see why anyone would want to invest in Gilts with negative real yields. Other than as a protection policy against a further deterioration of the eurozone, there is no real attraction. The return has gone but there are still risks. Gilts have a very asymmetric risk-return profile," says Wickham.
This Gilt-aversion is exactly what the Bank of England intended. By driving down yields it has forced investors into other asset classes, including corporate bonds. That has lowered borrowing costs for companies and the broader economy. Equity markets have also rallied.
The FTSE100 index of UK blue chips was at 3,645 when the first round of QE was announced, it is now over 5,700. However, with many retail investors wary of stocks after a decade of negative returns, the biggest flows have been into corporate bonds, the best selling sector tracked by the UK Investment Management Association for eight consecutive months.
The assets of domestic bond funds have increased by 114%. Over half of those recently polled by the CFA Society of the UK now view corporate bonds as overvalued. "The distorting effects of QE are not just limited to government bonds. They are writ large at the macro level across markets," says Mark Connolly, director of fixed income at Scottish Widows Investment Partnership (SWIP) in Edinburgh, which manages $218bn.
Investor base rebalances
For the structure of the Gilt market, the biggest effect of QE has been the change in the composition of the investor base. There has traditionally been a large domestic bid from the pension fund and insurance company sectors. In 2003, when Stheeman joined the DMO, these domestic institutions owned two-thirds of the market.
UK pension funds have increased their allocation to fixed income from 30% in 2005 to 47% now. That increasing allocation to fixed income is driven by several intertwined and enduring trends. In 1979 92% of individuals covered by a private pension were in a defined benefit scheme, one linked to final salary. Now, where defined benefit pensions still exist in the private sector, only 19% of schemes are open to new members.
Closed defined benefit schemes have very different investment priorities. Their future liabilities are easier to model. Asset-liability matching has also become more attractive because of regulation. Before QE, falling interest rates and the large domestic bid for liability matching assets were the biggest drivers of yields. That bid resulted in a persistently inverted yield curve, which only disinverted in the wake of the collapse of Lehman Brothers in September 2008.
"I think a normal range for the 30 year Gilt given interest rates and where they are attractive to pension funds is between 3.75% and 5%. Three percent is well below the level at which most pension funds will use Gilts to hedge liabilities," says Insights Wickham.
Francis Todd, managing director and head of Gilt trading at Goldman Sachs, also reports LDI funds are on the sidelines. "The LDI firms used to be the biggest source of demand at the long end of the curve. Now a broader range of investors, including from overseas, are also active in long maturities," he says.
There will always be some domestic demand for Gilts when investors have sterling-based benchmarks and when risk aversion is high. "If you move away from Gilts you are taking basis risk," says SWIPs Connolly. "The matching asset does what it says. Though it is likely almost any well diversified portfolio of sterling credit will outperform Gilts, there is a risk that the hedge does not deliver. In this atmosphere of uncertainty that mismatch makes some investors very nervous."
But, the DMOs figures are clear. Domestic holding of Gilts as a percentage of the overall market has fallen to around one-third over the past decade. That also needs to be seen both in the context of more issuance and QE. Some £164bn of Gilt issuance in 2012-2013 implies net purchases of around £50bn from domestic insurance companies and pension funds. That is more than all the Gilts issued in 2004-2005.
GEMMs (Gilt-edged Market Makers) report the continued active participation of domestic investors in auctions and secondary markets. "The feedback we get from UK investors is consistent: the yields are too low," says Myles Clarke, co-head of global syndicate at RBS. "But when we open the books there is always appetite for these bonds from domestic buyers. We continue to sell large amounts of Gilts to a diverse range of UK institutional investors."
Given that QE happens in the secondary market, the Bank of England has to buy those Gilts from somebody. QE plus bank purchases for regulatory liquidity buffers are big sources of new demand, more than £400bn. Safe haven buyers and overseas official institutions are the other groups displacing global asset allocators and domestic institutions.
Premium product, deep market
"We would argue the best place in bond markets are issuers with strong balance sheets, which means select investment grade corporates and emerging market sovereigns," says Mike Amey, portfolio manager and managing director in the London office of Pimco, which manages $1.8tr. "The biggest reason to hold Gilts is that they will outperform if Europe disintegrates and those safe haven flows have been significant."
Some domestic investors are concerned that these overseas buyers may prove fickle friends. "I think there is a little bit of worry that if [European central bank president Mario] Draghi does stabilise the periphery through bond buying, you may see some reversal of the safe haven flows, both out of Gilts and sterling at the margin. Though I do think we are talking about the margins," says Andrew Milligan, head of strategy at Standard Life Investments in Edinburgh which has assets under management of $250bn.
According to the GEMMs, most of these new investors in Gilts are official institutions and sovereign wealth funds and their commitment to the market is likely to remain durable. Global foreign exchange reserves held by central banks have soared by $4tr since 2008. Though the International Monetary Funds Currency Composition of Official Foreign Exchange Reserves (COFER) database suggests only a modest rise in percentage allocation to sterling to just over 20%, one-fifth of $4tr is a lot of new money in the system.
At the same time there is shrinking pool of AAA and alternative assets. A tactical asset allocator might buy the short end of Spain in anticipation of ECB bond buying, but strategic reserve policy is much more biased toward the long term. "There is a lot of liquidity out there," says RBSs Clarke, "and investors dont have access to the same suite of eligible products, such as CDOs [collateralised debt obligations] and CLOs [collateralised loan obligations]. Since Draghis announcement, we have seen significant buying of Spanish paper and the yields have collapsed."
Relative valuations are also important. Though negative real yields across a large part of the curve make Gilts expensive in absolute terms and set against their history, the same is true of US Treasuries and German Bunds. For investors seeking liquid assets in which to put significant cash to work, these three markets remain the nominal bonds of choice. "There are sovereigns that you can think of as being AAA+, such as Norway. But Bunds still have a lower yield because there is liquidity," says Insights Wickham.
Even a downgrade looks unlikely to dent demand for Gilts. Investors point to the recent experience of the US, which was downgraded by Standard & Poors but has seen yields continue to fall. S&P confirmed the UKs rating and stable outlook in late July. QE is not a concern, according to Frank Gill, senior director of European sovereign ratings at S&P. "We do not methodologically have a problem with central banks expanding their balance sheets if they can do so without creating a rise in long term inflation. Our view is that central banks, which are not regulated, can operate with negative levels of capital for a considerable period of time," he says.
SWIPs Connolly thinks that any ratings downgrade would only prompt a fundamental re-evaluation of Gilts if the economic or political backdrop changed. "If there was a sharp decline in growth prospects or a prolonged period of stagflation that would not be positive. But the commitment to reducing the deficit is the central issue. Thats in the coalition agreement which means there would have to be a change of government," says Connolly. "The label put on the debt by the ratings agencies is less important."
The depth and proper functioning of the Gilt market is built on both the investor base and the institutional framework of government debt issuance. This is a hidden strength that market participants value, particularly the GEMMs which put their balance sheet on the line. Each March, after extensive consultations with HM Treasury, the DMO publishes the Debt and Reserves Management Report (DRMR).
The DRMR sets out the overall size of debt issuance for each financial year, the planned maturity structure and the proportion of conventional (nominal) and index-linked issues. The government has a financing requirement that needs to be met, but the broad parameters of issuance are based on an iterative, bottom-up process in which the art of the possible is always the primary consideration.
Once this annual remit is set out, the issuance of particular Gilts, the auction dates and announcement of any syndications are then agreed on a quarterly basis. This is agreed after formal meetings with a group of GEMMs and end investors, as well as a large number of less formal bilateral discussions. The precise amount of issuance is not agreed until just before the auction. "We like to able to fine-tune according to market conditions," says Stheeman. "But broadly I would like to think we have the balance between transparency and flexibility correct."
The operation of monetary policy has traditionally been far more opaque. The Nobel prize-winning economist Robert Solow famously compared central bankers as being like squid because, "they emit ink and then move away". But QE plays such an important part in the Gilt market, that the Bank of England has needed to mindful of how its policies impact on it and the governments debt management objectives. It seems that the Bank of England has heeded the importance of transparency.
The operational policy framework for buying Gilts by the Bank of England Asset Purchase Facility Fund Limited is codified in a Consolidated Market Notice. The latest was published in March 2012. This sets out which Gilts the bank will buy (no linkers, minimum maturity three years, outstanding issue size of at least £4bn) and the reverse auction process. It is similar to the annual remit of the DMO, but only applies to each round of QE.
Each Thursday at 4pm the Bank of England announces the size of the auctions the following week and which Gilts will be purchased. These details are confirmed at 9am on the day. "The transparency means that dealers know what to expect and can position accordingly," says RBSs Clarke. "Given the scope of QE and amount of issuance, the market has remained liquid and has functioned well."
Stheeman says that though debt management and monetary policy remain distinct, where co-ordination has been necessary open lines of communication have worked. He cites an example from the repo market in the early days of QE in 2009. The DMO noticed that the GEMMs were using its standing repo facility in greater numbers and settlement failures were increasing.
"That was not in the interest of the Gilt market, the repo market or the broader money markets," says Stheeman. It was decided that the DMO should be able to lend parts of the Bank of Englands QE portfolio if necessary. "The Gilt repo market has worked very well for a number of years, which is conducive to good liquidity" says Goldmans Todd. "We have seen more severe repo squeezes in most other sovereign debt markets."
For Stheeman it is the proper functioning of the Gilt market that enables the DMO to sell bonds. The way the market has coped with QE demonstrates the strength of the institutional framework, including the DMO itself, the GEMMs and primary dealer system. He is less concerned about distribution channels or even price. It is the DMOs remit to limit costs to the Treasury, subject to risk.
Stheeman does not believe that the rebalancing of the investor base toward overseas institutions will have any detrimental effect. "You could argue that the market was previously too skewed toward domestic pension funds and insurers. Currently the distribution of Gilts means there is no single dominant type of investor. The existence of a deep and liquid Gilt market is the biggest factor supporting demand. If one group started to sell, another would step in, though that may be at a different price," he says.
There may also have been fringe benefits from overseas official institutions stepping into Gilts. Domestic investors are overwhelmingly interested in buying long dated assets. This new overseas bid has allowed the DMO to deepen liquidity at the short and medium part of the curve. "I think the DMO has done an excellent job managing the shape of the curve and the maturity profile of the debt," says SWIPs Connolly.
The average maturity profile of Gilts is perhaps the biggest single point of differentiation for the UK government bond market. At more than 14 years it is more than double the average of the remaining G7 countries. That has been possible because of the depth of the domestic investor base. Says Connolly: "There is almost a symbiotic relationship. The DMO knows it can issue longer dated paper because there is demand. That has helped lengthen the maturity of the Gilt market and it has helped us to manage longer term liabilities in a more thoughtful manner."
Before the financial crisis it seemed inconceivable that short term funding markets could close to significant borrowers, either banks or sovereigns. But it is now a risk credit rating agencies take seriously and international investors are increasingly attuned to. Jake Lowery, global rates manager at $170bn of assets ING US Investment Management in Atlanta, says: "The US, UK and France have very similar debt-to-GDP ratios. But the debt refinancing risk in the US is higher. That also needs to be seen against the backdrop of the fiscal cliff."
The most significant uncertainty facing Gilts is QE. Its withdrawal will likely be a multi-stage process and if it is handled poorly it could disrupt the market. The first stage of withdrawing QE would be a slower pace of purchases and that is what is widely assumed will be announced in November.Because UK growth continues to disappoint (see The long and winding road, page 4) the debt-to-GDP ratio remains high and so will issuance. Someone will need to take up that slack and with yields unattractive to a large number of prospective investors there is likely to be a price adjustment. The current sell-off suggests this has already been anticipated.
When the music stops
A bigger challenge still will be when the Bank of England stops all Gilt buying. That could come sooner than generally thought. Amey at Pimco says the fate of QE rests on the state of the economy, but that the Bank of England recognises the marginal value of the policy is diminishing. "Im not sure they got the bang for their buck they hoped for, though it has been supportive for asset prices."
If QE stops in the interim, the Bank of England is unlikely to reinvest the proceeds from its existing stock of debt as it matures. That will be the first step on the road of unwinding the programme.
The questionable impact of more QE also explains why a new policy prescription, credit easing, through the Funding for Lending scheme and other initiatives, was rushed out over the summer. "I think Funding for Lending is well constructed. It could get the economy going. The people I speak to in the Treasury certainly hope it could make a substantive difference," says Standard Lifes Milligan.
A return to growth in the second half of the year would see economists rushing to upgrade forecasts for 2013. That would make the outlook for Gilts very different to the current consensus. QE would certainly stop. But running off the debt would still be the likely course, rather than withdrawing liquidity through reverse repos or other means. The growth picture would have to be much clearer, Gilt issuance lower and interest rates rising for that to
A longer term concern is inflation. "I dont think you need to be an extreme monetarist to worry about the potential effect of QE on inflation," says Milligan. "Our pension fund clients have to think long term and many are beginning to voice concerns about an inflation spike and potentially large capital losses from Gilts in a replay of the US in 1994."
For the time being withdrawal of QE, rising interest rates and resurgent inflation all seem a distant prospect. The European Central Banks decision to buy the bonds of Italy and Spain is for some investors yet another reason to favour Gilts. Christoph Kind, head of asset allocation at 16bn of assets Frankfurt Trust Asset Management certainly thinks so. "For me Gilts relative to Bunds are attractive. The ECB buying the periphery is bad for the core. I think it is possible that Gilts could trade through Bunds at certain points on the curve."Kind is swift to add that this has more to do with the large structural bid from the Bank of England for Gilts and the likely reversal of some safe-haven flows into Bunds given the implicit liability of German taxpayers for peripheral debt. It is not a comment on the relative strength of the UK economy. But if the DMOs Stheeman has found the pace of bond issuance "extraordinary", the UK government must surely look at the deficit and then the cost of borrowing and find it little short of surreal. Both are a testimony to the Gilt market.