US TIPS vs Gold, Part 3: From Inflation to Real Rates

Summary

• TIPS-implied real rates are being driven lower by supply/demand imbalances, the Fed’s forward guidance on interest rates and balance-sheet policies, and inflation hedging fears.
• Furthermore, if the Fed can achieve its new framework of average inflation targeting, real rates, as implied by USTs after CPI adjustments, could decline to levels last seen in the 1970s.
• Gold has a strong inverse correlation with real rate movements, so another prolonged period of financial repression and negative real rates supports gold.
• Over the medium- to long-term, gold is a beneficiary of waning discretionary demand for USTs, especially from foreign investors concerned about FIAT currencies generally.
• Despite possible periods of temporary dollar strength ahead as the global economy calibrates to the post-Covid-19 world, over the longer term there is a case for the dollar likely declining by 20-30% or more in this new cycle. This too is gold friendly.

Market Implications

• Long term – bearish the dollar, especially versus gold; real rates may make new lows too, and not only in dollars.

As they say, three time’s a charm, so we went ahead and produced one more piece on TIPS vs gold. We had originally planned only a two-part series, with the specific objective of reviewing how well these two assets perform as inflation hedges and on a relative basis versus other real assets (Part 1 and Part 2 here). However, given the strong response and interest in our prior pieces, we thought it would be helpful to review the formation and evolution of real rates and how that relates to historical gold prices.

Before we explore the connection between real rates and gold, one needs to understand what exactly real rates are and identify the determining forces that shape them. Put simply, real rates are the level of interest rates that are the remainder that one earns after having factored in inflation on risk-free (i.e. government) bonds. One can also assume that, over extended periods of time, long-term real rates should align closely with the real potential growth rate (i.e. real GDP) of an economy as well. These macro fundamental identities are what link real rates to the real economy.

That said, just like any other assets, TIPS, which are the purest form of real rates, can be impacted by market forces such as supply/demand imbalances and investors’ search for yield. As such, in reality, real rates as captured by pricing from the market can often diverge from the textbook definition.

How Non-Discretionary Bond Buyers and Fed Action in Both TIPS and Treasuries Impact Real Rates

In a Macro Hive piece that co-author George Goncalves wrote in May, he discussed how foreign investors have been losing their dominant position in the UST market over the last five years and how domestic investors have been playing a bigger role. The combination of non-discretionary buyers (largely due to regulatory reasons) and Fed QE in TIPS and Treasuries has impacted term premia out the curve, helping to keep the overall level of yields lower, including real rates.

In Chart 1 above, we see that non-discretionary investors’ holdings of government bonds as a percent of total UST debt outstanding (where we group banks, pensions and insurance companies) have remained a stable share ever since the GFC. That is impressive, given we have lived in a world of declining yields and an ever-enlarging US government bond market. It was only during the Fed’s large-scale QE3 UST buying (large for back then, that is) that we witnessed a brief dip in the percentage of UST total holdings of these non-discretionary investors. Even in the post-Covid-19 massive Fed buying spree of USTs and TIPS, these stable buyers of UST offer a stark contrast relative to money market funds and mutual fund activity, which have generally diverged from trend, including recently. This steady support for the UST market is one reason why interest rates, both nominal and real, are stable.

Now, it’s also worth noting that recent Fed purchases of TIPS have been much larger versus previous QE operations. Under the objective of supporting market functioning above all else (where they are not calling it QE per se for Treasuries, TIPS and MBS), the Fed has actually more than doubled its holdings of TIPS (currently on track to break $300bn) since the start of the Covid-19 crisis.

In a world now desensitised to billions (and trillions are becoming the norm too), the nearly $300bn number is huge when seen as a percentage of the total TIPS market. As we highlight in Chart 2, the Fed is closing in on 20% of the TIPS debt outstanding. Therefore, it is unsurprising that TIPS quickly rallied and that the illiquidity premium, which had temporarily formed during the few weeks at the start of the Covid-19 market dislocation, vanished alongside real rates dropping like a stone. It is amazing what can happen when the Fed buys that much paper and so quickly. The last time the Fed made a larger commitment to TIPS was back during QE2. Let’s remember it only took to move their TIPS holdings over 10% of the market to move things. What is interesting, but expected, is that in both instances, when the Fed buys more TIPS than prior open market campaigns, gold rallies hard.

A final important point worth making here is that there is another, more subtle force at work in the market that impacts the valuations for longer-dated Treasuries, both in nominals and in TIPS space. While non-discretionary buyers may comprise a large portion of the market, directional trends are far more likely to be driven by those taking a ‘view’, that is, discretionary investors or speculators. Many years ago, co-author John Butler developed a concept he termed the ‘Duration Paradox’, an analytical framework for understanding the relationships between interest rate level, curve slope, curvature, convexity and the volatility surface as short nominal rates approach zero.

In brief, other factors equal, as rates approach zero, discretionary investors with a bullish bias can no longer express their views in shorter maturities, and so they are ‘forced’ out the curve. Bearish investors, by contrast, can express their views as before, as there is no ‘upper bound’ equivalent to the zero ‘lower bound’ (ZLB) (and even here we are being generous given that negative rates seen in some parts of Europe and in Japan are still close to the ZLB and thus feel asymmetrical). As such, while bullish investors in the markets must move out the curve, bearish investors can remain where they are, and the net effect is a bullish flattening in excess of that which would occur were the distribution of outcomes more symmetric – i.e. were short rates at a more normal level thus allowing for a potentially large decline.

This compression of the curve as short-end nominal rates go to even lower lows works for both the nominal and real rate curve and by extension compresses term premia generally too. The ‘Duration Paradox’ implies myriad other distortions. But, for our purposes here, the point is that the overall effects serve to reinforce the direct actions by the Fed and non-discretionary investors to drive real yields even lower.

Is Gold Just the Mirror Image of TIPS Real Yields?

On the surface and as Chart 3 below demonstrates, it does look like gold has benefited more from the movement in TIPS real rates versus pure changes in inflation expectations. That is not to say that gold and TIPS always move inversely to one another. For the record, this chart is a rolling annual change in the level of 10yr US real rates (in basis point terms) versus the rolling annual price change in gold.

During the euphoria of 2011, the prior peak in gold prices, there was a temporary divergence from the behaviour displayed in the TIPS real rates complex. And there were uncorrelated moves once the Fed started hiking in 2015. However, something interesting happened as we got towards the end of that hiking campaign, as if the gold market knew the next thing coming was not a Fed on perma-hold at higher rates, but instead that they would need to pivot and eventually do some easing measures. The Powell Pivot marked both the top in real rates and the start of a multi-year gold market rally.

Negative Rates Can Go to Levels Never Imagined, Well Not Since the 1970s at Least

So, in modern history the ebbs and flows of real rates have had an undeniable influence on the directionality of gold. But is that also the case further back in time? The answer is a resounding yes. The theoretical basis for this relationship is one of opportunity cost – the better investors are compensated by real returns on risk-free debt, the less attractive the alternative: risk-free gold. But when a central bank acts to suppress real rates, the relative attractiveness of gold rises. Co-author John Butler wrote about this relationship in some detail in Part II of his Gold Series earlier this year.

The TIPS market has only been around since the 1990s. As such, to demonstrate this relationship, we constructed a proxy for real rates using the 10yr UST nominal rate and subtracted out either the rolling year-over-year change in headline or core CPI. As we highlight in Chart 3, in the 1970s real rates by these measures went even further negative versus recently achieved levels. In both instances that real rates declined for a sustained period below negative 2%, the gold price rose sharply.

If the Fed is really committed to its new average inflation target and keeps rates on hold for years to come, and if we do end up with higher actual inflation in the process, that will create conditions not dissimilar from the 1970s. In other words, a stagflation environment, with gold likely to outperform.

At Some Point Gold Stops Being Just a Real Rate Play and Is a View on FIAT Currencies

While international investors may now comprise a smaller share of the Treasury market than in the past, they naturally remain very important. That said, with the notable exception of central banks, foreign investors normally have significant discretion as to whether to purchase Treasuries or other foreign assets instead. These decisions are naturally driven as much or more by the foreign exchange outlook as that for interest rates. Some foreign investors hedge their holding of foreign securities back into their base currency; others do not. But on balance, foreign asset purchases are not 100% hedged. And to the extent they are not, investors are taking a view on relative currency movements.

As co-author John Butler wrote back in a two-part series in June (Part 1 and Part 2) and updated this month, the US dollar has probably peaked for this cycle. In brief, the dollar remains strong in trade-weighted terms, yet both the US fiscal and monetary authorities are acting in ways which imply that the dollar should begin to feel the weight of its historically outsized current-account deficit. This in cumulative terms is an order of magnitude larger as a percentage of GDP than at the beginning of prior periods of material, sustained dollar weakness. Indeed, once the global economy re-calibrates post Covid-19, a case can be made that the dollar is likely to decline by 20-30% or more in this cycle.

Foreign investors sharing even a mild version of this view may therefore continue to slow their purchases of Treasuries. However, they may also see growing risks in their domestic bond markets, where in some cases yields are already near or even below zero. Returning briefly to the ‘opportunity cost’ argument for gold above, such investors are increasingly likely to see gold not only as an attractive alternative to a weaker dollar but also to their own domestic bond markets. The overall effect of such a demand shift on gold, a market with a relatively fixed supply, could be dramatic to say the least.