The premature end of low bond yields?
Updated: Apr 2, 2020
There are thoughts that the long-term bond rally (e.g. 30yr US Treasury yield above) might be over - what does the data and recent policy indicate?
Reasons for the end of the bull run include huge supply to pay for the Covid19 aide packages (see https://www.thegreyfirehorse.com/covid19-packages).
Despite the large 380% increase in US federal debt over the last 10 years, debt servicing costs have only increased 180%
Lower interest rates and the refinancing of maturing high coupon bonds into lower coupon bonds have contributed to lowering the average interest rate on outstanding debt.
The small Interest Rate increases in 2018 had a large impact on the annual interest cost, depicting how the large debt stock in now very sensitive to changes in interest rates
Over the last 10 years, the Fed, like the ECB and BOE, has prevented debt servicing to become a budgetary issue. There is no reason to expect a change to this policy. Indeed, the BOJ has been doing this for 20 years.
This suggests that expecting a rise in yields is premature, just like the widow maker of trying to short JGBs.
The price to pay for countries who use direct intervention to manipulate interest rates to keep interest rate costs low will be a weaker currency. This is classic EM economics, with precedent of this occurring in DM in previous centuries.
Gold is made attractive by a line up of unattractive currencies all being debased. The USD should remain bid for the reasons suggested last week – global deleveraging required dollars to repay $ debt.
The numerous announcements of fiscal aide packages, accompanied by the sharp moves higher in longer term government bond yields, has raised the debate on what is the fair value for government bond yields. Talk of bond bears taking charge, the onset of inflation, deleveraging of existing held outstanding debt and the prospects of a deluge of issuance have all contributed to make some market participants call the end of the multi-decade bond market. To this end, I have set out the hard data to allow for an informed decision on whether it’s all over for long end bonds.
My previous blog (https://www.thegreyfirehorse.com/post/real-rates-and-market-confidence-1) outlined the potential supply picture, and how it would be financed. The market should expect literally trillions of new issuance if all the government aide pledges come to fruition (I have outlined these on a new separate page Covid19 Aide). If anything, this should impact the SHORT END of the yield curve where the increased bond supply will be overwhelmingly issued. This is what happened when bond supply increased in the years after the Great Financial Crisis (GFC), when Trump started his fiscal largess and during the Italian Bond crisis (it took a year for the Tesoro to be able to issue long dated paper again, so it had to rely on issuing short dated maturities to refinance its' debt). So the market is correct to expect a large increase in bond issuance, but it is likely to be predominantly on the shorter end of the yield curve. But this will not necessarily impact outright yield levels but the value of US Treasuries against the implied future interest rate curve (forward interest rate swap curve). Indeed, the result should be a cheapening of USTs vs swaps, as has already started happening since the debt ballooned post GFC. Remember that during the Clinton budget positive years, where there was large scale negative supply, nominal yields rose, but US Treasuries outperformed swaps as USTs became more scarce.
My investing framework is based on a deep understanding of how assets and liabilities cohabit in a fiat currency world. It has Fisher’s Debt Deflationary Theory at its core (link at bottom) and is built by following credit trends. One of the key stages of a credit cycle is understanding when a “state of over indebtedness” exists, which then makes that economy very prone to a deleveraging shock with an Irving Fisher 9-step vicious spiral to asset price deflation. But it is not all doom and gloom, because as Fisher stated, if the “fall of assets is interfered with by reflation or otherwise” then the system is held together, and asset prices can “reflate”. That is exactly what QE has been in the last 10 years.
A state of over indebtedness (Cause) brings some very important consequences (Effect) for a government’s debt servicing, budget deficits and leverage. Understanding these effects is necessary to have a view on long end yields. Like in the last blog, I will demonstrate data for the US Treasury market, but the dynamics are the same in most other countries with high Sovereign Debt/GDP ratios.
US Treasury Debt Growth
Charts below show the tremendous growth in outstanding debt, the relatively small proportion of outstanding 30y Bonds (14% of total debt), the importance of foreign investors who now own 41% of the debt, and how the FED has had a varying but relatively small part to play in absorbing the national debt (pre crisis 17% of total, max 20%, end Feb20 14%).
US Debt Servicing Costs
Despite the vast increase in outstanding debt, servicing costs have not become an issue for the US Treasury. There are various factors at play that determine the annual interest cost:
1) Amount outstanding: Obviously the larger the debt, the higher the interest cost if interest rates remain unchanged.
2) Market Interest Rate: Similarly, the lower the interest rate, the lower the interest cost for the same amount of debt outstanding.
3) The shape of the curve and where debt is issued. A steep curve will mean higher costs if long term debt is issued and vice versa when short end debt is issued. Flat curves make issuance costs indifferent on the maturity of the debt issued.
4) Refinancing of maturing debt at new market rates. If interest rates decline significantly, then old debt with higher interest rate costs will be refinanced at lower interest rates, reducing the total interest cost.
The left chart below shows the average annual interest rate that the US Treasury paid on all its’ outstanding marketable debt over the last 19 years, and the yearly cost of that interest in $bn. Whilst the debt has significantly increased by 370% since 2007, the interest cost has only increased by 180%. The chart on the right shows the average interest rate for different bill and bond maturities, which clearly depicts the sensitivity of Bills, FRNs and Notes to prevailing interest rates, whilst the average interest rate of longer term Bonds is slowly declining as old Bonds mature and get refinanced into new much lower interest rate Bonds (e.g. in May $7.5bn of the 30yr T 8.75% May20 will refinance into a 1.3% couponed 30yr).
But note the little kink upwards in 2018 in the charts above and the impact on interest payments – it was a very timely reminder of how sensitive the debt stock had become to changes in interest rates. The mere 200bp increase in interest costs led to an almost $ 100bn increase in the ANNUAL interest cost. The above left chart is reproduced below for Bills and FRNs (left) and Coupon (USTs and Tips) paper (right).
The chart below summarises the average interest rate (%), the interest cost as a percentage of GDP (%), and the growth of Debt and GDP (both $ bn). This is the clear evidence that the debt remains serviceable if rates are low, and that an excessive amount of the budget will not be destined to interest costs (let alone to repay it!), which would have a negative impact on growth.
It’s simple to play around with the interest rate and outstanding debt data (given that outstanding debt is about to significantly increase) and see how a relatively small increase in interest rates would cause an immediate deterioration of the budget deficit. This is why I expect the US Treasury and the FED to do everything at their disposal to avoid yields rising, including debt monetisation through expanding the FED's balance sheet to buy USTs. This would permanently rake up the debt to avoid the private sector from having to absorb it at the expense of holding private sector assets. If they do not do this and the private sector has to buy the debt, then private sector assets would undoubtedly suffer.
What I have described above is familiar territory for participants of the JGB market. Together with shorting Bunds, shorting JGBs has been one of the big widow makers over the last 20 years, and I suspect that in the near term the same will be the case for those who short long maturities of other developed markets’ yield curves. The political and economic ramifications of rising yields are just far to big – anything will be done to ensure they do not rise.
This is the reason why in my investing framework I expect QE to always be ramped up when debt issuance increases. The larger the issuance, the larger the size of the program, width in scope and immediacy of action. Last week’s blog recommended to think the unthinkable as the sizes of outstanding debt and potential new issuance were so large that they would have to be matched by large and very creative solutions – today’s FED announcement has proved this point. Expect more.
This should keep a cap on real yields from rising (see Friday’s post) and lead to a crystallising of the "Japanification" of the US interest rate curve. You could argue that this has already started, as the shape of the UST curve (green) to the swap curve (blue) is almost identical in both countries. The only difference is that Japanese long term interest rates are expected to be lower (30yr swaps are 0.2% in Japan vs 0.8% in the US). A final move down in long end US rates to those levels is probably the final juice left on the US curve, which is scant consolation for balanced funds or pension/insurance companies who need fixed income to counter equity or longevity liabilities.
The final comment is on the pressure valve through which all imbalances must flow – a country’s currency. The above description of a debt trap and the ensuing reflationary policy is not normally found in a Developed Market economy - it’s usually found in EM space. However debt monetisation has occurred many times in the past in developed economies – just not in the recent past. Furthermore, too many people have believed for one reason or another that “QE” was temporary in nature and not plain debt monetisation.
When an EM country issues too much debt, then tries to print its’ way out of its’ “state of over-indebtedness”, imbalances fly out through the pressure valve that is its’ currency, which devalues by a sufficient amount to render the country competitive and to attract foreign investors again.
The problem today for market participants is that everyone is printing money, so investors are left with an attractive line- up of currencies – probably why Gold has its allure. Once the market realises just how much money printing, gold should move the next leg up. After all, the proper definition of inflation is how many dollars are needed to purchase the same amount of a good or service. If the amount of Gold remains the same, and the amount of $s increases by $ 1trn plus, then the price of Gold will surely rise.
I believe the USD remains the least unattractive currency simply because as long as global deleveraging continues, a determined bid on the greenback will remain in place from non-US debtors who have borrowed in $s. Only once deleveraging slows and reverses should the $ lose its appeal – but I don’t expect that in the near future.
“The debt-deflation theory of great depressions” by Irving Fisher https://phare.univ-paris1.fr/fileadmin/PHARE/Irving_Fisher_1933.pdf)
US Debt Data
https://www.sifma.org/resources/research/us-marketable-treasury-issuance-outstanding-and-interest-rates/ https://ticdata.treasury.gov/Publish/mfh.txt https://www.treasurydirect.gov/govt/rates/pd/avg/avg.htm