It is with some degree of resignation that I herein offer my opinion of gold as an investment. I generally try not to add my voice to the cacophony that’s sometimes created around “hot topics”. That said, gold’s rise over the past months is hard to ignore, and several clients and friends of HIA have asked our opinion of the metal.
In this paper I intend to investigate the rationale (if, indeed, investment markets can be said to be rational) those currently investing in gold may be leaning upon and attempt to evaluate the data backing those claims. The proposed explanations and rationalizations for investment in gold are:
- Gold is a good hedge against inflation, current or expected
- Gold is a good hedge against deflation
- Gold is a good hedge against a collapsing US Dollar
- Gold is a good hedge against uncertain financial conditions
I’ll also discuss other reasons investors may be interested in gold, and note the strategy Harmonic has pursued.
A technical note: Since the suitability of an investment as a hedge is determined by how well it tracks the item it’s supposed to be hedging, I’ve run correlation analyses for gold versus each of the items above. The correlation graphs plot the historical values of gold and the item to be hedged, and the “best fit line” shows the general tendency over the time frame analyzed. I’ve provided “R-squared” as the measure of correlation; simply remember that an R2 of 0.000 means the two items are totally uncorrelated, while 1.000 means they’re perfectly correlated.
Review
Gold’s history as a store of value and currency is well-known and will not be recited here. Observing gold’s historical price changes offers a quick reminder that since 1973, when the US Dollar was removed from the “gold standard”, the metal has seen two major price rises – most recently within the past 5 years, which eclipsed that of the late 1970s and early 1980s.

Gold as a Current Inflation Hedge
Gold has been widely touted as a hedge against inflation. Indeed, one branch of economic orthodoxy suggests exactly that, and this explanation, like the others examined here, has been provided as an explanation for gold’s recent run-up.
Comparing the long-term data, however, leaves one less than sure about that claim. The upper chart in the exhibit below compares reported consumer inflation and the spot price of gold (that is, for immediate delivery rather than a futures contract) from early 1973, approximately the date the US Dollar was taken off the “gold standard”, to present. The correlation (reported as R2) between CPI and the price of gold over that term is rather low, at 0.385.

By contrast, the correlation between CPI and the gold price is substantially higher in recent periods. The lower graph shows the correlation since early 2001, around the time when the FOMC began lowering the Fed Funds rate to ostensibly blunt the effects of the bursting of the stock market bubble. The increased correlation between gold and CPI is clear, at almost 93% (0.926).
In recent years it appears that gold has become an excellent hedge for consumer inflation. One must wonder when observing the increased dispersion of the data points on the right side of the lower graph, however, whether this tendency will continue.
Gold as an Expected Inflation Hedge
By definition, the monthly Consumer Price Index reflects historical price data. It stands to reason that investors in gold may be more interested in hedging themselves against potential future price increases than historical ones.
Since Treasury Inflation-protected Securities (TIPS) are constructed so that changes in consumer prices are reflected in their principal value, the difference in yield between those bonds and nominal 10-year US Treasury notes is an approximation of the market’s expectation of inflation over the upcoming 10 years.

The top graph shows that, since the creation of TIPS in the late 1990s, there has been effectively no correlation between gold and expected inflation (R2 = 0.004). However, that correlation, like that of the CPI measured above, has increased dramatically in recent periods, as shown on the lower graph. Weekly data since the cut of the Federal Funds rate to 0% in mid-December 2008 shows a widely dispersed but much higher R2 of 0.508.
Gold as a Deflation Hedge
In comparison to the enormous volume of writings about gold’s efficacy, or lack thereof, as a hedge against inflation, there is precious little research on gold as a hedge against deflation. One probable reason is that extended deflationary periods are rare in the US, which makes quantitative analysis based on the domestic market difficult. In fact, the only relevant period during the 20th century would be the Great Depression, during which time the price of gold was fixed in the sense that the US$ was on the gold standard. However, it’s reasonable to infer that since deflation is corrosive to financial assets such as stocks, non-financial assets like commodities may perform relatively well. Their low correlation to other asset classes is, after all, the reason that Harmonic includes commodities in our clients’ portfolios.
The best modern subject for analysis might be the Japanese economy, which has suffered deflation from the early 1990s to date. However, since gold is priced in US Dollars, such an analysis is really only a backdoor way of evaluating Dollar/Yen exchange rates, which of course is not the subject of this paper.
Gold as a US Dollar Hedge
Probably more than any other argument that’s been put forth of late, the rising price of gold has been attributed to investors concerned about a “collapse” of the US Dollar (US$). Whether the US$ is in fact collapsing is an analysis for another paper; my investigation here is focused only on whether gold is a good hedge against currency debasement. The chart below, graphing the value since 1973 of the trade-weighted US$, shows both the long decline since May 2002 and the recent spike and “collapse”.

A look at the correlation between the US$ and gold shows, like those analyses correlating gold against the CPI and implied inflation, a relatively low long-term correlation that has risen in recent periods. The analysis presented below uses the same time periods as the one plotting gold against CPI presented above: The upper graph shows the R2 of the US$ versus spot gold at less than 25% (R2 = 0.249), but the R2 since early 2001 is around 67% (R2 = 0.674).

Gold as a Hedge Against Financial System Risk
The Bloomberg Financial Conditions Index evaluates the level of inputs from the money, stock and bond markets to measure stress in the financial system (if you’d like to have the underlying calculations of the Index, please let me know). The gold price shows a low correlation to the Index back to its earliest calculation in 1991 (R2 = 0.210); further, since the spike in the Index following Lehman Brothers’ collapse in mid-December 2007, the correlation has actually declined marginally (R2 = 0.147).

Interesting Observations
Comparing the normalized performance of CPI, expected inflation, the US$ and gold provides several interesting observations. Firstly, the US$ (red line) and expected inflation (yellow line) have fallen dramatically since 1997. Secondly, actual reported consumer inflation has risen steadily over the past 13 years (which also suggests another observation – that market participants, at least the ones buying TIPS, don’t do a very good job of forecasting inflation). Thirdly and most importantly for this analysis, the normalized price of gold rose dramatically beginning in mid-2005, which begs the question “why?”

Perhaps not coincidentally, the most popular and largest gold-focused exchange-traded fund, the S&P Gold Trust (ticker = GLD) began trading in November 2004.

As shown below, the correlation between the price of gold and the number of GLD shares outstanding is quite high, with an R2 of over 84% (0.842).

While of course correlation does not necessarily imply causation, I find it very interesting that the price of gold tracks so closely the rise in popularity of an investment vehicle through which individuals and institutional investors may invest in gold bullion without the well-known pitfalls, such as illiquidity and storage costs.
Harmonic Investment Advisors Exposure
Rather than invest directly in gold bullion, coins or futures contracts, HIA invests most clients’ funds in an exchange-traded fund (the iShares S&P GSCI Commodity-indexed Trust, ticker = GSG) that is constructed to mirror the performance of a basket of commodities, specifically the Goldman Sachs Commodity Index. The index, and therefore GSG, holds about 2.75% of its value in gold, so most of our clients have exposure to gold indirectly.
Conclusions
Quantitative evidence suggests many of the assertions being put forth to justify gold purchases are less than iron-clad. While it appears that correlations between gold and CPI, expected inflation and the US$ have increased in recent periods, there is a low correlation between gold and inflation, expected inflation, US$ and financial conditions over the longer-term periods examined. Whether gold is a good long-term hedge in a deflationary environment is inconclusive due to lack of quantitative evidence.
For shorter time spans, gold has been an increasingly accurate inflation hedge in recent years. Whether it continues to be so obviously remains to be seen.
However, the simplest explanation for the rise of gold price may be the correct one – gold’s rising because investors are buying more of it, outpacing supply. If this situation is the case, it clearly implies some degree of speculation – purchasing without a fundamental basis but rather on the belief that others will also find the asset of value and drive the price higher. In this case, that projection may have recently become a self-fulfilling prophecy.
However, I don’t believe gold is yet in a bubble in the traditional sense. According to Merrill Lynch, the gold price would have to climb to over $6000/oz to get the ratio of gold to stocks back to where it was in the bubble of the 70s. While it’s received attention from the popular press, there’s not (yet) a mania for gold. There are, however, increasing demand drivers in addition to the aforementioned ETFs and mutual funds, like central banks (note India’s recent large purchase, with China likely to be next) and retail demand (gold is now being sold over the counter at the Harrods department store in London, cash-for-gold kiosks are springing up at local shopping malls, etc.)
While it is difficult to quantitatively analyze, it appears to me that gold is a good crisis hedge. Obviously it’s impossible to reach a consensus on what constitutes a crisis, but observing the periods during which gold has risen I note that extremes in the gold price coincide with periods of high economic uncertainty – particularly the late 1970s and over the past year.
Also, no investment is the panacea to all financial problems. As shown in the chart below, gold (green line) has periodically underperformed other asset classes like stocks (white line = S&P500), bonds (red line = Barclays Capital Aggregate Index), and real estate (yellow line = Dow Jones US Real Estate Securities index) over long periods of time.

Further, I rhetorically ask the true “gold bugs” who may be purchasing in expectation of a “Mad Max” sort of apocalyptic future: Wouldn’t productive assets like water, fuel and land be better stores of value than a metal one cannot eat, grow or burn for heat? (this is a conclusion, by the way, of Barton Biggs recent book “Wealth War & Wisdom”) In that sense, gold is perhaps the ultimate fiat currency – it has value only because we say it does.
Nov 25th
Weekly Economic Insight, 23 November – 27 November
Author: Kenn Lamson
Comments: 0
RELEASE
PERIOD
ACTUAL
EXPECTED (consensus)
LAST
HIA COMMENT
(leading, coincident, or lagging indicator)
October
6.10M
5.70M
5.57M
3Q09 preliminary
2.8%
2.8%
3.5%
September
146.5
NA
146.0
October
-0.6%
0.5%
2.0%
October (MoM)
0.7%
0.5%
-0.6%
November
67.4
67.0
70.6
October
430K
410K
405K
Positively, the estimate of housing stock outstanding continued to fall; the measure now stands at 7 months. Government stimulus is clearly underpinning the housing market, where buyers rushed to “get under the wire” to receive the first-time homebuyers credit, initially scheduled to expire in November. The housing market’s also being supported in a less obvious way, through the Federal Reserve purchase of mortgage-backed securities, a key mechanism for providing liquidity to lenders and keeping mortgage interest rates down. Those purchases, however, expand the Fed’s balance sheet and exacerbate longer-term inflationary concerns.
This week’s release is our second look at 3Q GDP, which was revised downward as expected. If the obvious stimulative effect of “cash-for-clunkers”, the first-time homebuyer’s tax incentive, and other measures were removed, the growth rate would clearly be significantly lower. The downward revisions to consumer and business spending are troubling, since (1) business investment is a precursor of stability in the manufacturing sector, widely touted as likely to be a key driver of the US economy out of recession, and (2) as we have consistently observed, since the US consumer represents about 70% of economic activity, demand from that sector is critical.
Orders for new long-lived goods rebounded lower from a surge in September. While a useful anticipatory measure of manufacturing sector activity, the data on durable goods orders is a very volatile series.
As expected consumer spending data was boosted by the “cash-for-clunkers” program. Personal savings declined slightly, to 4.4% from 4.6% in September.
Within the U of M Consumer Sentiment Index, the measure of current conditions, which reflects Americans’ perceptions of their own finances and whether it is a good time to buy big ticket items such as cars and homes, fell to 68.8 from 73.7, which was the highest in a year. The index of expectations six months from now, which more closely projects the direction of consumer spending, decreased to 66.5 from 68.6.
The margin of error for New Home Sales 6.2% month-over-month increase was 17.6%, so as usual the initial estimate should be taken with a grain of salt. Also, the gains were not distributed evenly across the US, as the South saw most of the uptick. As with existing home sales, the spike in new home sales can in part be attributed to the recently-extended and expanded first-time homebuyer tax credit. The declining volume of new homes for sale reflects both that purchase activity has increased and that, as indicated in the recently discussed construction statistics, homebuilders have slowed the pace at which they’ve added to the housing stock.