In our 2019 Q2 forecast we expected bond yields to remain stable until the end of the year. However, the yield on ten-year US Treasuries fell by 110 basis points in the three months to August, surprising many including Fathom.
Naively extrapolating the chart below would lead to the conclusion that yields will be deep into negative territory within the next decade, at least amongst advanced economies. A key factor determining the yields is the policy rate, which passes through to the economy in the form of deposit rates. For example, Switzerland has a negative policy rate, set by the central bank, and a negative-yielding ten-year government bond. Could it also have negative deposit rates for consumers? Conventional wisdom would say no. People could just resort to storing cash under a mattress and enjoy a return of zero percent. Yet UBS Switzerland is preparing to charge 0.75% on large deposits. If zero (or slightly below) is not the lower bound, what is?
Refresh the chart in your browser | Edit chart in Datastream
A rough answer to this question comes from thinking about the practicalities of using cash instead of electronic forms of payment.
Cash is not always easy to store. Storing cash under a mattress may be feasible for amounts up to £1000, but few are likely to be comfortable with doing this with larger sums. Most people would think that an amount like £50,000 would need to be stored in a safe, which if large enough would have to be stored in another secure location, which might even warrant employing people to guard the safe — a bit like a bank. A conservative estimate suggests that the cost of storage of cash outside of bank deposits would be around 25 basis points of the value of the cash holding per year.
Neither is cash convenient. Receiving or spending large amounts of money, such as being paid a salary or buying a car, would involve carrying thousands of pounds from one place to another. The literature on this suggests between zero and 200 basis points as the cost of convenience of holding physical cash.
Lastly, if one were to store large quantities of cash at home, it is likely that insurers would catch on and raise their premia. Another conservative estimate would point to 25 basis points being the cost of insurance.
Between those three factors, it is not unreasonable to argue that consumers would be prepared to pay up to 250 basis points — an interest rate of -2.5% — in order to enjoy the benefits offered by a deposit account. To go beyond that would require a change in the status quo for cash. It would have to be taxed or abolished. The government could introduce a 1% tax on the use of ATMs. Or cash could be ‘stamped’ with a date, which could determine by how much its value has diminished since it was printed.
In the limit however, taxing cash will encourage people to look for alternatives. The first option would be foreign currency, especially ‘hard’ currencies that already function as numeraire currencies in many markets. So, a ‘cash tax’ could only work if coordinated internationally, but not every country will be willing to do so. Switching to precious metals as a medium of exchange is also unlikely to work. Not only does it come with issues such as counterfeiting, adulteration, and shifts in supply of the chosen metal, but it would also need to be stored somewhere. Banks could be obliged by the government to tax these coin deposits. To avoid this, consumers would have to arrange their own storage facilities which would probably be more inconvenient than storing cash. Cryptocurrencies can also be ruled out. While the supply of any individual cryptocurrency such as Bitcoin may be fixed, the supply of cryptocurrencies in general is not. There could be infinitely many types of cryptocurrency, and with no intrinsic value or any sovereign backer to differentiate between, there is no way to value one against the other. 
Refresh the chart in your browser | Edit chart in Datastream
Want more charts and analysis? Access a pre-built library of charts built by Fathom Consulting via Datastream Chartbook in Eikon.
This does not mean that cash will always be king. The recent growth in contactless payment systems suggests that physical mediums of exchange might soon be obsolete. If the government proposed withdrawing cash from circulation, as a prelude to introducing deeply negative interest rates, there is a risk that would induce a bank run. The ideal scenario is one where the use of cash has already diminished to almost nothing, and cash is then withdrawn from circulation without any fanfare.
Even in the presence of physical cash, some longer-term yields are already negative, and are being weighed down by numerous factors. The spread between the ten-year government bond yield and the ten-year overnight interest swap captures some of these effects, such as the appetite for the liquidity that is available in government bonds as opposed to long-dated OIS, and distortions arising from the use of QE, which would not affect OIS as much as government bonds. The spread is negative in Germany and Japan, close to zero in the UK, and in very low positive territory in the US.
Refresh the chart in your browser | Edit chart in Datastream
But in our view, the main reason for exceptionally low government bond yields is a low or negative term premium. It reflects the insurance that ‘safe’ bonds like USTs or Bunds offer against a 2008/09-style recession. In that scenario equities and other risk assets will fall, government bonds will rally and USTs, JGBs and Bunds — the ultimate safe-haven assets — will rally the most. That would be a result of private appetite for those assets, cuts in policy rates, and increased central bank purchases of government debt.
In current circumstances, the attractive insurance property of government debt is amplified by high debt-to-GDP ratios, which are associated with low growth, inflation, policy rates, and yields. High leverage ratios create a significant tail risk of a global financial crisis in the minds of investors, possibly exceeding the scale of 2008/09. Our analysis shows that there is evidence in the data of debt weighing down on firms’ investment decisions and contributing negatively to the term premium on government bonds.
Our conclusion is that it is time to take negative yields seriously. Deeply negative yields are not in Fathom’s central case, but deposit rates could in principle fall as far as -250 basis points, and long-term government bond yields to -500 basis points. If cash were rendered obsolete or abolished, then those lower bounds would shift down substantially in both cases. The implications of such rates at all horizons would be far-reaching. Models that value assets based on a discounted stream of future earnings would need to be rethought: what would be the appropriate equity price if the denominator in our dividend discount model were negative? What is the funding level of a pension scheme whose stream of future liabilities is discounted negatively? The world of profoundly negative yields is no longer such a distant prospect that it can be safely ignored.
The charts in this article have been created using Chartbook on Datastream. The Chartbook, created and maintained by Fathom Consulting, is a library of over 9000 charts, containing up-to-date macro and financial market data for over 170 countries. Whether it is a particular topic, country or variable you are interested in charting, the Chartbook has everything you need. Simply type search ‘cbook’ into your Eikon search bar or click the ‘Chartbook’ tab on Datastream to find out more.
__________________________________________________________________________________
Datastream
Financial time series database which allows you to identify and examine trends, generate and test ideas and develop view points on the market.
Refinitiv offers the world’s most comprehensive historical database for numerical macroeconomic and cross-asset financial data which started in the 1950s and has grown into an indispensable resource for financial professionals. Find out more.
