We know that Sterling has historically traded like a risk-on currency that tends to appreciate in periods of trending markets and consolidate sharply in high-volatility regime.
Figure 1 shows the monthly average performance of the most liquid currencies relative to the dollar when the VIX rises above 20 in the past 30 years. As expected, the yen is the currency that benefits the most when price volatility rises, averaging 45bps in monthly returns. On the other hand, the pound has averaged -30bps in monthly returns when the VIX was high.
This was confirmed during the March 2020 panic as GBP was sold aggressively during that month with Cable reaching a low of 1.14 (down from 1.32 in early March) before starting to recover gradually (lowest level since 1985).
Figure 2 shows an interesting relationship between GBPUSD and mega-cap growth stocks since 2020 (FANG+ stocks); Sterling has significantly recovered in the past 17 months, up nearly 20% against the US Dollar. However, the momentum on Cable has halted in recent months as risky assets have shown some signs of ‘fatigue’ amid rising uncertainty over a range of risk factors (i.e. Delta variant, falling growth expectations…).
Nominal yields on long-term government bonds have been constantly trending lower in the past 35 years, mainly driven by the lower global economic growth and the convenience yield for safety and liquidity. In addition, policy responses from both governments and central banks following the Great Financial Crisis have also contributed to the downtrend in LT interest rates globally, driving the term premium (investors’ compensation for holding interest rate risk) to negative levels.
Figure 1 represents the history of interest rates since the beginning of the 12th century up to today. The data source comes from Homer and Sylla’s book A History of Interest Rates (2005), which reviews interest rate trends in the major economies over four millennia of economic history. For the last 150 years, we compute a GDP-weighted average interest rate times series using a sample of 17 countries (Global LT interest Rate), using data from Jorda et al. paper The Rate of Return on Everything (2017).
One interesting observation was that in the past millennium, nominal interest rates have mainly traded below 10 percent, with two exceptions:
· The first one is during the Spanish Netherlands period, which is the name for the Habsburg Netherlands ruled by the Spanish branch of the Habsburg. In the 16th century, Antwerp was one of the most important financial centres in the world, dominating international markets in sugar, spices and textiles. During a significant part of this century, the Exchange at Antwerp had dominated European transactions in bills of exchange and other credit instruments such as demand notes, deposit certificates and the bonds of states and towns (all short-term debts). Between 1508 and 1570 (when Antwerp defaulted on its debt), the interest rates were from loans by various bankers to the Spanish Netherlands government, Charles V and the city of Antwerp. We can notice that interest rates surged to 20% in the mid-1520s, and the bankers’ family, the Fuggers, made emergency loans to Charles V at annual rates as high as 24 to 52 per cent.
· The second period of high interest rates was during the Great Inflation of the 1970s, which was marked by a sustained trend in inflation in the US that affected the rest of the World as well. A number of factors were associated to the rise in inflation during that decade: oil price shock following the 1973 embargo, speculation, avaricious union leaders and bad monetary policy management. As a consequence, long-term interest rates started to surge around the world, reaching a high of 14 per cent on average in 1981.
Interest rates globally are sitting at their lowest level in history, with the aggregate currently trading at around 0.5 per cent (probably trading between 1%-1.5% if we include some major EM economies to our times series of LT interest rates). As uncertainty has been constantly reaching new all-time highs (especially in the past year following the Covid19 shock) and growth expectations have been lowered accordingly, global yields have decreased dramatically since the last quarter of 2018 and broke below the lows reached in the late 16th century during the Republic of Genoa. At that time, Genoa became the banker for the Spanish Crown, a role previously held by German and Dutch bankers. The Bank of St. George was issuing placements or perpetual bonds called luoghi, which paid deferred dividends on the amount of taxes collected, after subtracting payment of the expenses of the bank.
The explanation of the persistence of very low real interest rates in advanced economies is a contentious issue. While many studies, including those by central bank research departments, stress secular forces such as demographics, over-indebtedness and dispersed income distributions – factors that emerged before the Great Financial Crisis – others find an important role for monetary policy in the determination of interest rates. On the secular view, a potential mean-reversion in long-term real interest rates must be driven by a global economic force such as a sustained period of above-trend economic growth. On the latter view, the monetary policy regime is endogenous to the real interest rate and the unconventional policies of the past decade are part of the explanation for low rates.
In the past few months, we have seen that China Total Social Financing (TSF), a broader measure of credit and liquidity, has been falling sharply with the annual change in TSF 12-month sum down from over 10tr CNY in October 2020 to nearly 0 in May 2021. As a result, investors’ concern has been growing as they have been questioning if the rally we have observed in global asset prices can continue in the coming 6 to 12 months without Chinese help.
In the past cycle , periods of contraction in Chinese liquidity were usually associated with a fall in both domestic and international asset prices. This chart shows the striking co-movement between the annual change in China Tech stocks (CQQQ ETF) and the annual change in China TSF 12M sum. China Tech stocks are down over 25% since their mid-February highs; according to this chart, China tech stocks are expected to continue to consolidate in the short term.
Even though some analysts have compared the 2020 rebound in equities to the 1930 ‘hope’ phase following the 1929 crash, we think that this year has shown some strong similarities with the 1998 / 1999 period. While tech stocks were experiencing strong inflows in the second half of the 1990s amid the dotcom boom, the Nasdaq suddenly fell by 30% in the third quarter of 1998, before starting to reach new highs and surging by over 120% in the following year.
This year, tech companies’ valuations are up 90% in the past 9 months following their dip reached on March 23rd and seem on their way to reach new all-time highs in the coming months as another 3 trillion USD is expected to reach markets in the coming year.
This chart shows some strong co-movements between the Nasdaq index in 1996 – 2000 and in August 2018 – December 2020. Even though market sentiment has reached extreme levels, the bullish trend in mega-cap growth stocks could easily continue for another year amid the surge in liquidity coming from central banks to support the economies and finance the high costs of lockdowns.
In the past few years, a significant amount of economists and practitioners have warned of a potential hard landing in the Australian housing market, as property prices have been growing at unsustainable rates with first-home buyers having difficulties saving a significant deposit to get a foothold in the market. According to the Australian Bureau of Statistics, the total value of residential property in Australia is now exceeding 7 trillion USD, by far the economy’s largest asset. As there are no ‘vehicles’ to short the Australian housing market as during the US subprime crisis, two alternative ways to short the property market was through either going short the Australian Dollar or short the banks. Prior the Covid19 crisis, banks’ mortgages were equivalent to approximately 80% of the country’ GDP, with most of them piled into the top 4 banks (Commonwealth, WestPac, ANZ and NAB).
Even though house prices were starting to decline significantly in 2018 and the beginning of 2019, with investors speculating that it was the start of the ‘hard landing’, the reversal in the global stance of monetary policy (from quantitative tightening to quantitative easing) combined with the surge in Chinese liquidity have generated strong support for the Australian property market in the past year. This chart shows an interesting co-movement between China excess liquidity (6M lead), which we compute as the difference between real M1 money growth and industrial production, and the Australian housing market. It seems that the downside risk in the Aussie property market should remain limited as money growth keeps accelerating in China.
Since the start of the year, we saw that the dramatic liquidity injections from central banks to prevent the economies from falling into a global deflationary depression has led to a sharp depreciation of most of the currencies, especially against assets with limited supply such as Gold, Silver and Bitcoin. The liquidity also generated a strong rebound in risky assets such as equities, reinforcing the trend on the mega-cap growth stocks (FAAMN companies). The NYSE FANG+ index, which provides exposure to 10 of today’s highly-traded tech giants, is now trading nearly 50% higher than its February peak, which has massively contributed to the recovery in the SP500.
In the past few months, we have noticed an interesting observation: it seems that the Fed (and other central banks) interventions have led to one ‘global trade’ as a significant amount of assets have strongly co-moved together in 2020. This chart shows the strong relationship between Bitcoin prices and the FANG+ index; a few weeks ago, cheaper equities due to the rise in uncertainty over US elections and the lack of stimulus has led to also lower Bitcoin prices. Are ‘Bitcoin bulls’ really hedged against a sudden reversal in equities?
The past few months have been marked by a significant depreciation of the US Dollar relative to major currencies. The massive liquidity injections from the Fed to desperately avoid the country from falling into a deflationary depression combined with the risk of a second ‘strict’ lockdown in some southern states (Sun Belt) have left the greenback vulnerable, with some investors speculating that the ‘Bear Market of the USD’ has eventually begun. The USD index is now down 1.7% year-to-date, following two years of positive momentum, and currently trades at its lowest level since September 2018. The US Dollar has significantly weaken against the major currencies and particularly the Euro as EU leaders have recently shown improving signs of coordination and the ‘better management’ in the Covid19 crisis will surely be reflected in the real growth differential between the two economies, which is one of the key drivers of currencies over the long run.
However, as the USD index is heavily weighted on the Euro (57.6%), it may show a misleading picture of the current state of the US Dollar. If we look at a broader measure of the US Dollar – Nominal Effective Exchange Rate (NEER), which measures performance of the USD against a broader rage of currencies weighted according to the value of trade with the domestic country – we can notice that USD NEER is up 3.1% YtD and still trades higher than its early March level. The chart below shows the interesting ‘divergence’ we observe since the start of the crisis. At this stage, the USD NEER index has given up most of its March gains, but is clearly not showing any signs of massive weakness coming ahead. Hence, we will need to see more downside on the USD NEER to confirm that a LT bear market on the USD has eventually started.
As more and more regions in developed economies have been put under a dramatic total lockdown amid growing concerns over Covid-19, central banks have started to cut rates aggressively in order to avoid a complete market meltdown. We saw in the previous week that both the Fed and the BoE held emergency meetings and cut rates by 50bps, the most since the Great Financial Crisis, benefiting from their positive benchmark interest rate to act faster than the rest of central banks. Economies already experiencing a NIRP policy (i.e. Sweden, Euro area) will probably implement or expand asset-purchase programmes in order to fight against a significant economic shock and therefore implicitly reduce their ‘shadow rate’, a rate first introduced by Fischer Black (1995) that can measure the effects of QE, to lower levels.
However, it is important to note that a significant reduction in benchmark policy rates globally has been associated with sharp equity sell-offs. This chart shows that in the previous two downturns, the GDP-weighted G10 policy rate was cut by approximately 4 percent and coincided with a global equity sell-off of 45% to 55%. Are we set for a similar story in 2020?
Chart. G10 policy rate vs. World equities (source: Eikon Reuters)
In the past two years, the elevated economic and political uncertainty in addition to the lagged effect of quantitative tightening have significantly weakened growth expectations and as a consequence increased demand for safe assets such as the US Dollar and US Treasuries. While the situation seems to have improved slightly in the past 6 months on the back of aggressive rate cuts from central banks globally, business surveys are still pricing in further deterioration in the US economy. For instance, the ISM manufacturing PMI hit a low of 47.2 in December 2019, diverging significantly from the 50-percent threshold that separates growth from contraction. CEO confidence also dropped to its lowest level in a decade and is currently pricing a much higher probability of recession than other popular indicators. On the other hand, consumer confidence indicators have remained strong in the US as consumption remains solid (real PCE expenditure has been averaging 2.5% in the past few quarters).
How long can that divergence persist until US consumer sentiment starts to fade away? This great chart shows that the CEO confidence survey has acted as a good 12-month leading indicator of consumer confidence (University of Michigan) since 1980. We can notice that top executives in the US are currently pricing a significant deterioration in consumer sentiment for the next 12 months to come. However, some divergences occurred in the past, particularly in the late 1990s when CEO confidence started to fall drastically in 1998 and 1999, but consumer confidence was constantly rising during that period mainly due to the tremendous rise in equities that was inflating household wealth. Even though we are concerned about the deterioration of those business surveys, we may continue to see a divergence within the next twelve months between business and consumer confidence surveys as equities keep reaching new all-time highs and interest rates remain ‘too low’ relative to the current pace of nominal GDP growth in the US.
Chart. US CEO Confidence (12M Lead) vs. Consumer Confidence (Source: Eikon Reuters)
Empirical researchers have demonstrated that gold has had many drivers over the past few decades, but has been mainly influenced by interest rates, inflation trends, the US Dollar, stock prices and central banks reserve policies. Baur and McDermott (2010) also shows that the precious metal plays the of a safe ‘zero-beta’ asset in periods of market stress and equity selloffs. For instance, in the last quarter of 2018, US equities (SP500) fell by 14% while the price of gold in US Dollars was up 7.6%. In the short run, participants usually look at the co-movement between gold price and real interest rate (TIPS) to define a fair value of the precious metal (gold price rises when real yields fall and vice versa).
However, gold has shown a stronger relationship with another variable in recent years: the amount of negative-yielding debt around the world. This chart shows us the striking co-movement between the two times series. After oscillating around USD 8 trillion between the beginning of 2016 and the end of 2018, the amount of negative-yielding debt doubled to nearly USD 17 trillion in the first half of 2019 amid political uncertainty and concerns over global growth, levitating gold prices from $1,280 to $1,525. However, we have noticed that investors’ concern has eased in the past two months, normalising global yields (to the upside), increasing the US 2Y10Y yield curve back to 25bps after turning negative in the end of August, therefore reducing preference for ‘safe’ assets such as bonds. The amount of debt yielding below 0% has dropped significantly since the end of August to USD 11.6 trillion this week, dragging down gold prices to $1,460. We think that market participants have overreacted to the global growth slowdown in the first half of the year and that the rise in leading indicators we have observed in the past three months (i.e. global manufacturing PMI) will continue to push preference for risk-on assets. The amount of negative-yielding debt could easily come back to its 2016-2018 8-trillion-dollar average in the following months, hence emphasising the downward pressure on gold prices. It looks like gold is set to retest the $1,350 – $1,400 support zone in the short run (which used to be its resistance zone before the 2019 rally).
Chart. Gold price (in USD) vs. amount of negative-yielding debt (tr USD) – Source: Bloomberg, Eikon Reuters.