Gold Glitters Out Of The 'Shadow'

Summary

• To obtain reliable forecasts of the price of gold, we propose to move the focus from a traditional "real rate view" to an overall monetary-policy-stance perspective.
• Recent contributions in applied monetary policy analysis have shown the usefulness of shadow, or implicit, rate indicators of the stance of monetary policy.
• The shadow short rate is equal to the policy rate in non-lower-bound/conventional monetary-policy environments, but it can freely evolve to negative values in lower-bound/unconventional environments.
• When we purge the data of the effects of the strong U.S. dollar, 2014-2017 turns out to have been an exceptional period in which gold appreciated while real interest rates climbed and monetary policy went through a tightening cycle.
• Gold prices can be expected to resume on an even stronger appreciation path because since early 2019, shadow short-term rate indicators have signaled a significant monetary policy easing in the U.S. and worldwide.

Editor's note: Seeking Alpha is proud to welcome Massimo Guidolin as a new contributor. It's easy to become a Seeking Alpha contributor and earn money for your best investment ideas. Active contributors also get free access to SA PREMIUM. Click here to find out more »

by Monia Magnani and Massimo Guidolin

In recent weeks, a flurry of news concerning the stance of monetary policy, both in the US and around the world, has impacted the price of gold and of its exchange-traded vehicles, such as SPDR Gold Trust ETF (NYSEARCA:GLD), causing a surge in volatility. In this article, we argue that over the medium term, this news will keep contributing to the creation of a bullish environment for gold. The extent of such a favorable background can be fully appreciated only by looking at overall, overarching measures of the stance of monetary policy, such as those based on the notion of shadow policy rates.

Hard to formulate strategies based on unstable drivers

Even though most commentators have found reasons that these developments have impressed a bullish tone to the gold market, there are two obstacles to concluding that gold should be expected to move higher in the medium term (six- to 18-month horizon).

First, there are a host of additional gold-price drivers that it seems has been missed by the discussions on central-bank actions. For instance, it has been argued that the tendency of the US federal budget deficit to spiral out of control is a harbinger of future accommodative monetary policies and hence of inflation in the long run. This may be triggering bullish trades in the gold market. In an interesting article, Jason Tillberg has noticed that real wages, deflated by commodity prices, have been declining while gold has recently appreciated - and that, because of this, we should not expect gold to keep shining because it has become already expensive given the tamed dynamics in US wages.

Second, even if we just focus on the relationships among gold, the US dollar external value, real interest rates, and the expectations on future (unconventional) monetary policies, what we should truly expect remains shrouded in thick fog for a simple reason: The correlations linking these factors to gold prices have shown to be unstable and often difficult to predict. Our goal is to explore this dimension and to show a new indicator of the monetary policy stance that encompasses both observable rates and measures reflecting quantitative easing. That indicator may help clarify one fact that has gone unnoticed thus far: 2014-2017 has been an exceptional period, and going forward gold bugs ought to resume their bullishness when monetary policy eases further, as has been occurring.

Netting out the effects of the movements in the US dollar

To illustrate the difficulty at identifying any stable relationships between the price of gold and well-recognized drivers such as the USD exchange rate and real interest rates, we have taken a first step towards simplifying the problem. To purge gold dynamics of US dollar influence, as a first step we have derived a spot gold price index in which the effects of USD movements are netted out.

As a first approximation, one can read the resulting USD-adjusted spot gold price in the following chart as the international value of gold when the effects of global monetary policies on the exchange rates are removed from the data. We need to remember this step later on, because our analysis can only lead to predictions concerning the price action of gold, given any forecasts we may have about the US dollar.

Figure 1: Purging Gold Prices from US Dollar Effects. Source: elaborations by the authors.

Figure 1 is clear: netting out the contribution of the US dollar from spot gold price dynamics does not make a first-order difference, with the exception of the 2014-2015 period, when the strong dollar would have justified a higher price of bullion compared with what we actually saw. So a long US dollar/short gold strategy would have paid off handsomely during those two years. However, around 2016, in spite of a wedge created by the anomalous 2014-2015 period, the two series resumed moving together. During 2018, glittered regardless of whether the effects of a weakening US dollar are taken into account. Moreover, over our 2002-2019 sample, the adjusted gold price has reached unprecedented peaks: Can this be expected to go on for much longer?

Be aware of the unstable correlations

At this point, many commentators will know exactly where to look for the culprit of the recent history of USD-adjusted gold prices: real interest rates, and especially US real rates. Their point is simply that gold is an unproductive asset, a mere reserve of value. When the real, expected profitability of the productive uses of wealth increases - because of business-cycle effects or as a result of policy interventions - the price of gold climbs. We, therefore, use data on the implicit (breakeven) inflation-protected (hence, real) yield from 10-year TIPS to obtain a market measure of the forward-looking, real rate. Roughly speaking, the following plot confirms that this standard, off-the-shelf intuition keeps applying to recent history.

Figure 2: USD-Adjusted Gold Prices and the US 10-year Real Interest Rate. Source: elaborations by the authors.

Visibly, high real rates, like in 2002-2008 when they ranged between 1.5% and 2.7% per year, are accompanied by low gold prices, while zero or negative rates (like between 2011 and 2013 and in the late summer of 2019) tend to bring about surging gold valuations. Indeed, the correlation between these two series (a measure of how deviations from their means associate over time) is -0.85.

However, the devil is in the details. In this case, we might say that the devil jumped out of the chart and into our laps when we magnified the picture. On a closer scrutiny, it is clear that during the financial crisis real rates gyrated and eventually declined while the price of gold kept creeping up. At the time, in an atmosphere of panic and with banks bailed out by a series of emergency measures, this was probably unsurprising. Even during the 2014-2016 period, the real rate simply oscillated within a narrow range, between 0% and 0.8%, while the price of gold gradually moved up. In 2018, real rates moved up temporarily while we all know that the price action of gold was bullish. Simple statistics confirm these facts. Between 2002 and 2007, the correlation between the US real rate and USD-adjusted gold was a positive 0.47; during the financial crisis (2008-2010), it declined to -0.43 which remains well below the full-sample estimate. Between 2011 and 2019 the correlation was negative but moved towards zero (-0.21). What gives?

The following figure frames the same series in the form of a scatter plot in which the real rate should "explain" gold prices. While the general, inverse relationship between the two series is not in doubt (see the linear fit in red, which returns a correlation of -0.85), a closer inspection reveals the presence of "patches" of data, occurring in correspondence to either high (in excess of 1.8%) or negative real rates, in which the relationship is approximately one of direct proportionality, when higher real rates may justify gold bugs bullishness.

Figure 3: The Relationship between US Real Rates and USD-Adjusted Gold Prices. The plot shows a kernel regression (that at each point over-weights the data close to the point and under-weights the data that are more distant from it) in green, that turns flat or increasing when rates are negative (upper left portion) or exceed 1.8% (bottom, right portion). Source: elaborations by the authors.

Extracting useful information from shadow rates

What we propose is to move our focus, and also to yield more reliable forecasts of the price of gold, from the traditional "real rate view" to an overall monetary-policy stance point of view. Our thesis is that gold prices are affected by the perception of current and future monetary policy as a whole, especially when we net out the effects of monetary policy that go through the USD exchange rate. Since the crisis, monetary policy has had trouble producing a strong effect on both real interest rates and on other more traditional measures of policy stance (for instance, the quantity of money, M2). Therefore we propose to quit identifying the policy with its outcomes, i.e., real interest rates changes, to measure instead the magnitude of the policy impulse. The reason is that under the unconventional strategies required by sticky inflationary expectations, interest rates no longer matter. As we argue below, such a novel perspective delivers one unexpected insight: We may be on the verge of a return to the so-called post-financial crisis "new normal" after a failed flirt with the classical, pre-financial crisis "old normal".

Recent academic research has emphasized the usefulness of shadow (implicit) rate indicators of the posture of policymaking. In this article, we resort to Leo Krippner's shadow short rate, for short SSR. The SSR is the shortest maturity rate from an estimated shadow yield curve. It is basically equal to the policy interest rate in non-lower-bound/conventional policy environments (e.g., August 2008), but the SSR can freely evolve to negative values in lower bound/unconventional environments (e.g., in July 2011) to indicate an overall stance of monetary policy that is more accommodative than a near-zero rate alone. The following chart compares the dynamics of (adjusted) gold prices with the SSR, here interpreted as an indicator of the overall monetary policy.

Figure 4: US Dollar-Adjusted Gold Prices and the US Shadow Short-Term Rate. Source: elaborations by the authors.

Visibly, even though the US never experienced negative short-term rates, SSR did turn negative between 2008 and 2015 as a result of the wave of quantitative easing measures adopted by the Fed. As one would expect, SSR is negatively correlated with the gold price: even though SSR is not a real rate, in an environment with sticky and depressed inflation expectations as the US has been in since 2009, variations in SSR are highly correlated with the implied real costs of borrowing. In fact, the correlation between USD-adjusted gold prices and SSR is -0.50 over the full sample. However, such a correlation is not as extreme and informative as the -0.85 reported in Figure 2. Moreover, a visual inspection of Figure 4 also shows that the correlation between gold prices and SSR has been highly unstable and in particular that between 2014 and 2017 gold prices and SSR grew hand in hand. In fact, while between 2002 and 2013 the correlation had been as extreme as -0.83, it turned high and positive between 2014 and 2017. One may conclude that our attempt at explaining gold prices with SSR has been unfruitful.

Figure 5: The Relationship over Time between the US Shadow Rate and USD-Adjusted Gold Prices. Source: elaborations by the authors.

On the contrary, one last chart is insightful: the inverse relationship between gold prices and SSR, taken as an overall indicator of monetary policy, is stable and reliable in our sample, with the exception of a specific sub-sample, 2014-2017. In fact, starting in early 2018, the correlation turns negative again and appears to be back to around -0.8. The point then becomes: what was so special about the 48 months between 2014 and 2017?

Our view is that an odd stretch just ended - one in which gold prices could move up as a reflection of a monetary policy that was struggling to normalize itself even amid the adoption of tightening measures. In light of the recent steps adopted by the ECB to re-open the season of QE and the debate on the opportunity of negative rates and the resumption of an "organic growth" of the Fed's balance sheet, going forward gold may appreciate only if real rates were to decline and if monetary policy to become expansionary again. In other words, the attempt by central bankers to return to a normal policy regime of positive rates - whether failed is not for us to judge - did take gold off the rails for the duration of almost an entire business cycle, causing downward pressures that are currently fading.

Even though we need to be mindful of the fact that most of our predictions concern spot gold prices after US dollar effects are netted out, most experts have recently been bearish on the outlook of the US dollar (see, e.g., Cliff Droke's Seeking Alpha article), also because of the pressure exercised by the declining interest rates. Given the strong, negative correlation between the USD and gold that we have reported, this represents a further reason to be bullish on gold. Moreover, an argument based on the existence of a 2014-2017 shift in correlations estimated from time series data suffers from the peril that new and unforeseen shifts may occur in the future. Finally, the reliability of investment signals based on the SSR remains unproven, and serious backtesting work will be needed to support any actionable guidance.

Our investment thesis

When we purge the data of the effects of the strong US dollar, 2014-2017 turns out to have been an exceptional period in which gold appreciated while real interest rates climbed and monetary policy went through a tightening cycle. This may have reflected a deep lack of trust by market participants in the attempt to normalize worldwide monetary policies with a return to positive real interest rates and the phasing out of quantitative easing. Since 2018, the global business cycle has deteriorated. This has required a new accommodative stance by policymakers who are currently poised to lower real interest rates. If the correlations from the pre-2014 sample were to prevail again, these measures forecast that gold will soon be back on an even stronger appreciation path.

Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.