Speculators in gold are playing with fire

Overview

In a recent article (Why investors needn’t – and presently shouldn’t – own gold, 7 April) I wrote: “why should investors own gold? Its advocates have long cited a handful of attributes which justify its purchase.” Unlike many others, who apparently act in response to anecdotes, rumours, idle opinions or others’ research, Leithner & Company disregards gossip – and collects its own data, conducts its own analyses and draws its own conclusions. Our analysis of gold ignored miners (which muddy bullion’s waters with company-related risk) and answered four crucial questions:

1. Is gold a reliable, long-term hedge against consumer price inflation? Does it withstand inflation’s effects? Does its rate of return rise as inflation accelerates?
2. Is it a reliable hedge against short-term economic and financial uncertainty? Is it a “safe haven” whose value rises in response to actual geo-political tensions and fears of crises?
3. Does it reliably withstand recession? Is it a “safe haven” whose value jumps during economic slumps?
4. Does it reduce a diversified portfolio’s overall volatility? Increase its return?

Our analysis reaffirmed gold’s general status, at times of impending high inflation or recession, as a hedge and safe haven; however, it also concluded that its boosters typically overstate these attributes. Gold’s returns mitigate the effects of high (but not low or moderate) levels of inflation; they also lessen the effects of recession, but no better than 10-year Treasury bonds.

Six months after that first article, what’s changed? Firstly, its price (adjusted for CPI and denominated in $US) has risen more than 35%. Over the year to September, it vaulted more than 47%; that’s more than at any time in almost 20 years.

Secondly, ever fewer people, it seems, are now acquiring it primarily because they’re cautious, i.e., seek a prudent means to hedge against rising consumer price inflation, actual geo-political tensions and fears of crises, recession, etc. Instead, ever more people are apparently buying it because they’ve become aggressive – that is, confidently believe that its price will continue to skyrocket.

Holding gold, in short, is no longer a means of mitigating a portfolio’s risk of overall loss: instead, buying it has become a means of aggressive speculation, that is, an allegedly sure and easy way to make “fast money.”

“For some reason,” Warren Buffett told The New York Times Magazine (1 April 1990), “people take their cues from price action rather than from values. What doesn’t work is when you start doing things that you don’t understand or because they worked last week for somebody else. The dumbest reason in the world to buy a stock is because it’s going up.”

Buffett’s point also applies to gold. In this article, which extends and elaborates my previous one, I reiterate key facts which should (but probably won’t) disquiet today’s overconfident herd (see also Why you’re probably overconfident – and what you can do about it, 14 February 2022).

A Brief Preliminary

On 15 August 1971, U.S. President Richard Nixon “demonetised” gold; that is, he ended the Federal Reserve’s obligation to foreign central banks (since 1933, Americans hadn’t legally been able to own gold bullion; Nixon also lifted this prohibition) to exchange an ounce of gold for a specified quantity of $US, and vice versa. Since the “Nixon shock” (which also included freezes of wage and prices, and surcharges on imports) the U.S. Government hasn’t fixed the price of gold; instead, its price – as well as the $US’s rate of exchange with other currencies – has fluctuated in response to market forces.

Valid comparisons compare like with like – such as assets whose prices fluctuate freely. My analysis will therefore exclude gold’s fixed (in terms of $US) price before August 1971 and analyse its fluctuating price thereafter.

Similarly, I’ve computed and analysed the total (distribution or dividend as well as capital gain or loss) return of 10-year U.S. Treasury bonds and the Standard & Poor’s 500 Index. Hence all references to “bonds” mean 10-year Treasuries and “stocks” mean the S&P 500. I’ve expressed all returns as geometric means (that is, as compound annual growth rates (CAGRs)) rather than arithmetic means. Finally, all returns are net of the U.S. Consumer Price Index (that is, on a CPI-adjusted basis) and monetary amounts have been denominated in $US.

One reader of Why investors needn’t – and presently shouldn’t – own gold (7 April) objected: “exclusively looking at USD gold prices doesn’t make sense in my opinion if you’re talking to an Australian audience or managing Australian portfolios ... So currency is very relevant here.”

I replied: “yes, Australians assess their portfolios and wealth in $A, not in $US; and yes, currency is very relevant. Which is a big problem for you: it causes you greatly to overestimate gold’s return in CPI-adjusted $A. The reason is that the exchange rate of the $A has virtually halved since August 1971 – from $A1 = $US1.15 in that month to $A1 = $US0.62 in December 2024, and to $US0.60 or less in recent days.” (From December 2024 to September 2025, the $A has risen to ca. $US0.66.)

“In other words,” I replied, “the alleged ‘outperformance’ of gold (denominated) in $A as opposed to $US ... , which you imply, is mostly attributable to the depreciation of the $A since 1971. $A gold hasn’t outperformed $US gold: the $A has underperformed the $US.”

I concluded: “gold is a globally-traded commodity; hence arbitrageurs’ high-speed computers will quickly remove any currency-related price discrepancies. Net of the exchange rate, I suspect that any remaining discrepancy in ($A-denominated versus $US-denominated) CPI-adjusted figures is (largely) attributable to CPI’s higher rate of growth in Australia (CAGR of 4.99% per year since Aug1971) than in the U.S. (3.92%).”

Let’s Begin with a Half-Century of Context

Figure 1 plots gold’s price per ounce, in CPI-adjusted $US, since August 1971. Three long cycles are apparent. Firstly, during the 1970s it rose almost without interruption: from $339 in August 1971 to $2,598 in September 1980. Secondly, it collapsed to $498 in July 1999.

Figure 1: Gold’s Price per Ounce, Monthly, CPI-Adjusted $US, August 1971-September 2025

Anybody who claims that gold has always been a “store of value,” and that it invariably withstands the effects of inflation, has rocks in his head: from September 1980 to July 1999, its CPI-adjusted price plunged 81% and at a compound annual growth rate (CAGR) of -8.3% per year.

In July 1999, however, gold commenced a long comeback – to $3,747 in September 2025. That’s a total gain of 652% and a CAGR of 8.4% per year. Particularly noticeable is its skyrocketing price since late-2022. Notice as well, however, that only in September 2024 did its price reclaim the CPI-adjusted pinnacle it first climbed in September 1980. During these 44 years, it generated a total return of exactly 0.0%. That’s grossly deficient protection against consumer price inflation.

It’s clear in retrospect that in September 1980 gold was egregiously overvalued. Hence the crucial question: has it once again become grossly overpriced?

Figure 2 plots gold’s 12-month, CPI-adjusted rolling return since August 1972, that is, from August 1971 to August 1972, September 1971 to September 1972, ... and September 2024 to September 2025. In the 12 months to September 2025, its price zoomed 47.4%. That’s the biggest increase since the 53.7% in the 12 months to May 2006, and exceeds the increases during and immediately after the GFC; yet it pales in comparison to the massive rises (more than 160%) in 1980, and those of ca. 75% in the mid-1970s.

Figure 2: Gold’s Price, 12-Month Rolling Percentage Change, CPI-Adjusted $US, August 1972-September 2025

Suppose that in August 1971 you (or your forbears) purchased equal quantities ($100) of gold, 10-year U.S. Treasury bonds and a portfolio of stocks which perfectly mimicked the Standard & Poor’s 500 Index, and that since then you (or your heirs) have reinvested all dividends and payments of interest. Net of the Consumer Price Index, and ignoring holding costs, transactions fees, taxes, etc., how would your investments have waxed and waned each month to September 2025? Figure 3 plots the answers.

Figure 3: CPI-Adjusted Total Returns, Three Investments, Monthly, $US, August 1971-September 2025

Three of its results are most significant. Firstly, the investment of $100 in the S&P 500 Index in August 1971 grew to $3,592 in September 2025; net of consumer price inflation, that’s a compound annual growth rate (CAGR) of 6.9% per year over these 54 years. The investment in gold has grown to $1,105; that’s a CAGR of 4.5% per year; and the investment in the Treasury bonds has grown to $414; that’s a CAGR of 2.7% per year.

Over this single 54-year interval, gold has clearly outperformed Treasuries. Is it thereby reasonable to infer that over all long term (say, 20-year) periods that gold generally outperforms bonds? The answer is clearly “no.”

As I’ve already noted, over some very long intervals owners of gold have suffered massive losses. From January 1980 to April 2001, for example, it collapsed 83% and its CAGR was -8.0% per year for 21.25 years. And until the COVID-19 pandemic, bonds’ and gold’s 48-year CAGRs were indistinguishable.

Figure 3’s third result provides another caution: since October 2022, gold has skyrocketed 107%; however, before this spurt its CAGR from August 1971 to October 2022 was 3.3% per year – versus bonds’ 2.8%.

Figure 4 plots rolling 20-year total returns, expressed as CPI- adjusted CAGRs, of bonds, gold and stocks since August 1991. During the 1990s, gold’s returns collapsed deep into negative territory. Since the turn of the century, however, they’ve risen practically without interruption. As a result, its most recent CAGR (8.6% per year) is by far the highest on record.

Figure 4: CPI-Adjusted Total Returns (20-Year CAGRs), Three Asset Classes, Monthly, August 1991-September 2025

Bonds’ 20-year CPI-adjusted CAGRs, in contrast, haven’t merely been mostly falling since the turn of the century: since 2023 its most recent total long-term returns (less than 1% per year) have been by far the lowest on record.

Then, bonds and stocks outperformed gold; today – for the first time – gold’s 20-year CAGR matches stocks’, and both have greatly outperformed bonds.

Gold Has Certainly Boomed; Has It Become a Mania?

According to a wide variety of sources, ranging from popular (such as Investopedia) to academic (e.g., Robert Shiller, Irrational Exuberance, Princeton University Press, 3^rd ed., 2015), the signs of a mania are many and varied – and comprise three broad categories:

#1: If “It’s Different This Time,” It’s Likely a Mania

During a mania, prices don’t just surge: they attain levels which “fundamentals” and traditional methods of valuation can’t justify. That’s why, as the mania intensifies, enthusiasts increasingly ignore or deny basics and abandon venerable methods and standards of valuation. They do so because, they assert, “it’s different this time!”

At all times, speculators care little about basics and time-tested yardsticks; during manias, they increasingly disparage those who do (“why can’t they understand that it’s different this time?”). 

Even more than during less exuberant periods, during booms speculators obsess over short-term movements of price; and their expectations, boosted by recently booming prices, become ever more over-optimistic and overconfident.

#2: If You’re Idiot Brother-in-Law is Bragging, It’s Likely a Mania

During manias, growing numbers of people, most of them inexperienced speculators, enter ebullient market sectors – and conservative investors face rising pressure to become aggressive speculators. Driven by the fear of foregoing seemingly large, fast and easy profits (“FOMO”), they make ill-considered decisions. Charles Kindleberger, the historian of financial crises, expressed this key point tartly: “there is nothing so disturbing to one’s well-being and judgment as to see a friend get rich.” Physically as well as figuratively, they compete – that is, queue – to buy mania-driven assets.

Manias intensify fear and greed. As they advance, participants become increasingly confident – and dismissive of non-participants. Those who haven’t participated in the mania often feel envious of those who have – and increasingly fear that, apparently unlike the participants, they won’t “get rich.”

During a mania, people buy not because they’ve conducted their own rigorous analyses, have conservatively estimated an asset’s value and concluded that its value exceeds its price: they buy simply because many others are aggressively buying, and thereby boosting the manic asset’s price.

They’re oblivious to the fact that their behaviour triggers a feedback loop that violates common sense (not to mention the laws of micro-economics): as the asset’s price increases, demand from speculators rises, which further boosts prices, etc.

Towards the crest of a boom, a widespread (and, at extremes and as Robert Shiller and others have documented, irrational) sense of exuberance prevails, and speculators dismiss concerns about risk and valuation. The belief spreads that, insofar as these anointed assets are concerned, current market conditions are fundamentally different from those of the past, and therefore that previous rules of valuation no longer apply.

Above all, during a mania anyone who presents logic and evidence which questions – never mind rejects – the mania is regarded as naïve, foolish or even malign.

During a financial mania, brokers and funds managers happily feed the euphoria. They seek to attract “investors” (that is, speculators) with promises of allegedly quick, safe and high returns by “investing” in assets whose prices have been rising most rapidly, and aggressively issuing stock of companies in “hot” sectors.

These funds appeal to a pool of “investors” who’re often inexperienced and driven by the fear of foregoing apparently easy profits.

The performance of aggressive and speculative funds during a mania can create another feedback loop: those which generate the “hottest” returns attract the most (“hot”) money, which further increases their demand for “hot” assets, which further fuels the mania.

These “hot” funds reject traditional methods and standards of valuation; instead, they feed the momentum and sentiment which are boosting prices.

#3: If It’s a Mania, “Experts” and the Media Will Become Even More Irresponsible

“Experts” and journalists don’t merely participate in manias: they actively fan their flames. “Old rules don’t apply in world’s rush for gold” (The Australian, 22 October) is typical. “Critics are missing something,” it confidently opined. “The gold market has changed. The old rules don’t apply.”

“As gold continues to surge,” it continued, “strategists are struggling to keep up.” It cited one who has “just upgraded his gold forecasts again. His middle scenario now has gold at $US4427 an ounce in 2026. Its high scenario is $5,108.” (There’s no low scenario. That’s typical of the euphoria during a mania; if the forecast is sunshine forever, why consider the possibility of storms?)

Such “forecasts,” I suspect, are no more than rose-tinted guesses, i.e., rubbish which merely tells aggressive speculators what they want to hear.

Robert Shiller agrees. In Irrational Exuberance he wrote: “unknown to most investors is the troubling lack of credibility in the quality of research being done on the stock market (by “analysts” at major financial institutions) ... Some of this so-called research often seems no more rigorous than the reading of the tea leaves.”

On 22 October, The Australian’s markets editor credulously concluded: “the gold market has fundamentally changed. Understanding why requires looking beyond inflation to the deeper structural shifts in global finance and geopolitics.”

He quoted an official at the World Gold Council (which on this matter can hardly be regarded as dispassionate): “the fundamentals for the gold market still remain very, very strong, and ... will lead to ongoing support for the gold price going forward.” 

I’d be amazed if they’ve ever said anything else!

Three Crucial Facts Today’s Speculators Apparently Don’t Know about Gold

#1: Gold Is Generally High-Volatility and Low-Return

“You’ve got to be prepared,” said Warren Buffett at a Berkshire Hathaway AGM (YouTube’s video doesn’t provide a date), “when you buy a stock to have it go down 50% or more ... If you can’t handle (such plunges) psychologically, then you really shouldn’t own stocks because you’re going to buy and sell them at the wrong time.”

Buffett’s point applies even more to gold than it does to stocks: whether it’s over all 12-month, five-year, ten-year or twenty-year intervals since August 1971, the volatility (measured by its standard deviation) of gold’s CPI-adjusted return exceeds the 10-year Treasury’s and the S&P 500’s (Figure 5a).

Figure 5a: Average Standard Deviations, CPI-Adjusted Total Returns, Three Investments, August 1971-September 2025

Since 1971, gold has generally been a high-volatility, low-return asset; bonds, in contrast, have been low-volatility and medium-return; and equities have been high-volatility, high-return.

It’s also crucial to comprehend what today’s speculators in gold apparently don’t: generally since August 1971, whether on short-term (12-month), medium-term (five-year) or long-term (10-year or more) bases, gold has generally been a relatively poor-performing asset (Figure 5b).

Figure 5b: Average CAGRs, CPI-Adjusted Total Returns, Three Investments, August 1971-September 2025

Gold can hedge against consumer price inflation: it tends to generate higher returns when CPI’s rate of change is high and rising, and lower returns when it’s low and falling. Bonds, in contrast, tend to underperform when CPI’s rate of change is high and rising, and to outperform when it’s low and falling. One of gold’s crucial shortcomings is that CPI’s rate of increase has generally – and cumulatively greatly – decelerated since the early-1970s. Accordingly, as the time interval increases from short-term to long-term, gold’s average CAGR has fallen, bonds’ has risen and stocks’ has remained roughly constant.

Indeed, over all intervals since August 1971, gold’s average CAGR increasingly underperforms bonds’ as well as stocks’.

#2: The Short-Term (12-Month) Volatility of Gold’s Return Is Mostly Unpredictable

In Stop kidding yourself: Nobody can “time the market” (30 June) I demonstrated the short-term, CPI-adjusted total returns of markets (such as the S&P 500 Index, All Ordinaries Index, etc.) and stocks (such as BHP, etc.) are almost perfectly random – and therefore, under most circumstances unpredictable. As a result, and except under very rare conditions, which I specified, virtually nobody can consistently “time the market.”

Accordingly, unless you’re a perfectly prescient trader, which nobody ever has been, is or will be, “buy and hold” investment generates higher average short-term returns than speculation. Finally, short-term results compound over longer periods: hence investment (“time in the market”) handily beats speculation (“timing the market”).

In Forget next year’s commodity prices: focus on 2075’s (1 September) I extended this logic and evidence and concluded that, over very long intervals, short-term random fluctuations (“noise”) cancel one another – and ever clearer and steadier signals emerge.

Except under rare circumstances, market-timers and commodity speculators can’t consistently speculate successfully. That is, they can’t repeatedly and accurately anticipate short-term movements of a market’s level or a stock’s or commodity’s price, return, etc. That’s because these fluctuations reflect mere chance. Random movements are irregular variations around a mean. They have no identifiable cause and are trendless; hence they’re unpredictable. Successful speculations are merely lucky guesses; these “successes” are therefore ephemeral.

Market-timing fails because it ignores a fundamental, obvious and insuperable difficulty: stocks’ and markets’ returns are random; by definition, random phenomena are unforeseeable – and nobody can predict what’s inherently unpredictable.

A Quantile-Quantile (“QQ”) plot provides one means to assess the extent to which a variable follows a normal distribution. (A quantile is a value below which a given percentage of data falls.) A QQ plot compares the quantiles of the S&P 500’s 12-month returns to those of a normal distribution whose mean and standard deviation equal the returns’. If the returns are normally distributed, the points on the QQ plot will fall along a straight line. A perfectly normal distribution thus has an R² of 1.0.

Figure 6 shows that, except at the extremes, gold’s 12-month returns since August 1971 have been normally distributed. Its returns, in other words, are basically random – and thus essentially unpredictable.

Figure 6: Quantile-Quantile Plot, Gold’s CPI-Adjusted Total 12-Month Returns, August 1971-September 2025

The distribution does, however, have “fat tails.” Large gains and losses occur more frequently than they would if these returns were perfectly normally-distributed (over the same interval, the corresponding R² for bonds is 0.968 and for stocks is 0.985).

The same phenomenon which often destroys hedge funds – the underestimation of the probability of major loss – could also destroy today’s speculators in gold (for details, see How Warren Buffett has trounced “the world’s greatest hedge fund manager” 11 August).

#3: Gold’s Most Recent Short, Medium and Long-Term Returns Are Grossly Unrepresentative

Thus far, two of my key results, which parallel those in Forget next year’s commodity prices: focus on 2075’s (1 September), bear emphasis. Firstly, in the short term nobody can reliably predict commodities’ (including gold’s) prices; secondly, over these intervals, they occasionally boom and bust.

We’re now in a position to establish a third key result. On the one hand, over very long terms and like other commodities (iron ore, oil, etc.), gold’s CPI-adjusted CAGR fluctuates far less than its 12-month return. On the other, its most recent short, medium and long-term returns are grossly unrepresentative of these returns as a whole (Table 1).

It also demonstrates that the most recent returns over all intervals – and, more generally and by implication, prices since October 2022 – have been unduly high.

Table 1: Gold’s Short, Medium, Long and Very Long-Term CAGRs, CPI-Adjusted, August 1971-September 2025

Sharp-eyed readers will observe from Table 1 that, over the 12 months to September 2025, gold’s return is twice its standard deviation. Given that its returns since August 1971 have been approximately normally distributed, it’s reasonable to infer that its recent return is highly unusual. Specifically, an observation from a normal distribution is likely to exceed twice its standard deviation approximately 2.5% of the time.

We can infer that, in the 12 months to September 2026, a consecutive strong gain is highly unlikely, i.e., will occur with a probability of 0.025 × 0.025 = 0.06%. Instead, it’s reasonable to infer that its return over the next 12 months will be much closer to its mean since August 1971 (CPI-adjusted 6.3%).

Astute readers will also observe from Table 1 that, over intervals between five and 40 years, the mean CAGR’s standard deviation is as large as or larger than the CAGR. For intervals of 10 years, for example, the mean CAGR is 1.9% per year and its standard deviation is 6.5%. From the essentially normal (albeit with fat tails) distribution of these CAGRs we can infer that the probability is 95% that, over the next decade, the CAGR is 1.9% ± (2 × 6.5%), i.e., lies between -11.1% and 14.9%. It’s entirely possible, in other words, that over the next decade gold generates a negative CPI-adjusted return.

Indeed, losses are entirely possible over intervals as long as 40 years. As we’ve seen (Figure 1), this has occurred before. A recurrence won’t surprise prudent investors. It would, however, shock today’s aggressive (because they’re overconfident) speculators.

What If It’s NOT Different This Time?

Table 1 implies that losses – including severe losses – are possible. Indeed, if over the next five years the most recent CAGRs – which, remember, are grossly unrepresentative of their long-term averages – revert to their averages since August 1971, then mediocre results or outright losses are more than likely.

Take gold’s average price in September ($US3,697), and note that its average, CPI-adjusted five-year CAGR since August 1971 is 2.9% per year. Net of consumer price inflation, its imputed price in September 2030 is thus $3,697 × (1.029)⁵ = $4,265. That’s a CPI-adjusted total gain of ($4,265 - $3,697) ÷ $3,697 = 15% (Table 2).

What if over the next five years gold’s 10-year CAGR reverts to its historical average of 1.9%? In September 2020, gold’s CPI-adjusted price was $2,392. A 10-year CAGR of 1.9% gives us a price of $2,392 × (1.019)¹⁰ = $2,887 in September 2030.

Compared to today’s price, that’s a total loss of 22% over the next five years, and a CAGR of -4.8% per year, and so on for the other intervals.

Table 2: Gold’s Imputed Price in September 2030

If over the next five years gold’s price conforms to its average 50-year CAGR of 2.5% per year, owners of gold (or buyers at current prices) will reap a bonanza. That’s because in the 12 months to September 1980 gold’s price leapt 69%, its CPI-adjusted price reached $2,598, and $2,598 × (1.025)⁵⁰ = $8,929. That’s a total gain of 142% and a CAGR of more than 19% per year.

This imputed return is a matter of the luck of the draw, i.e., at its starting point gold scaled one of its highest-ever CPI-adjusted prices. Overall, the odds are unattractive: the five-year CAGR generates a modest return, and reversions to 10, 20, 30 and 40-year CAGR generate grievous losses.

Investment and speculation are ultimately matters of odds. The odds typically favour long-term investors – and virtually always punish speculators. Gold’s present odds are highly unattractive to those who plan to buy it (or hold what they already own): in five of the six scenarios in Table 2, the prospective CAGR is mediocre or negative.

Why Hold Gold If Merely Moderate Inflation Is on the Horizon?

“Is ‘three’ the new ’two’?” editorialised The Wall Street Journal (“Getting Used to 3% Inflation”) on 24 October. According to data released that day by the Bureau of Labor Statistics, consumer prices in the U.S. increased ca. 3.0% year-on-year in September. WSJ noted: “no one in Washington seems bothered that this remains well above the Federal Reserve’s 2% inflation target.”

It added: “there’s always some excuse or explanation that politicians and Wall Street offer to say this is no big deal. One month it was healthcare costs, another month shelter, and so on. This month the blame goes to energy prices, which rose 1.5% from August to September. But at some point you have to admit all these add up to a persistent inflation problem. At a 3% inflation rate, the value of a dollar today would be 73.74 cents in 10 years.”

WSJ concludes: “for all the controversy over (President) Trump’s attempts to meddle with the Fed, he and (Chairman) Powell both seem to want lower interest rates. It makes us wonder if 3% is Washington’s new de facto inflation target. We doubt this is what anyone voted for in 2024.”

I doubt that CPI’s modest prospective rise can sustain gold’s recent meteoric rise; it hasn’t since August 1971 (see in particular Tables 2a and 2b in Why investors needn’t – and presently shouldn’t – own gold, 7 April).

Monthly since January 1981, the Federal Reserve Bank of Cleveland has estimated the expected rate of consumer price inflation in the U.S. over the next 12 months and five, ten, 20 and 30 years. Its estimates derive from a model which incorporates Treasury yields, inflation data, inflation swaps, and survey-based measures of expected consumer price inflation.

Figure 7: Consumer Price Inflation, Actual and 12-Month Forward Predictions, January 1983-September 2026

I’ve matched the Cleveland Fed’s 12-month forward estimates of consumer price inflation to the CPI’s actual annualised rate of change. For example, the two observations for January 1983 comprise (1) the Cleveland Fed’s estimate (which was formulated 12 months previously) and (2) the actual figure (based upon data released in ca. April 1983), and so on for succeeding months. Figure 7 plots the results.

The 12-month forward estimates have averaged 2.8%, and the actual 12-month rate has averaged 2.9%; accordingly, the average error is 0.1%, that is, the average estimate has slightly underestimated the average actual rate. Since September 2023, the estimates have averaged 2.7%.

Since the 1980s, the estimates haven’t anticipated sudden accelerations and decelerations (or outright decreases) of CPI; above all, they missed the sharp downward spikes in the immediate wake of the GFC and the sudden huge upsurge immediately after COVID-19.

The Cleveland Fed’s latest data (September 2025) estimate the rate of consumer price inflation in the 12 months to September 2026; we therefore lack actual inflation data to match to the most recent estimates. This gap becomes much longer when we consider the estimates of consumer price inflation for the next five years (Figure 8).

To fill it, I’ve assumed that henceforth CPI will increase at the rate of 0.25% per month, i.e., at a compound rate of 3.0% per year. I then incorporated this assumption into an estimate of CPI’s five-year CAGR. I’ve labeled these substitute estimates “Extrapolation to September 2030.” If my assumption is reasonable (compared to estimates for the next 12 months it slightly “overshoots”), over the next five years CPI’s actual rate of increase will sharply decelerate to the Cleveland Fed’s current five-year forward estimate.

Figure 8: Consumer Price Inflation, Actual and Five-Year Forward Predictions (CAGRs), January 1987-September 2030

The crucial question thus arises: why hold gold if no more than moderate inflation lurks on short-term (12-month) and medium-term (five-year) horizons? At its present price, I see little upside – and, in light of Table 2, significant downside.

Why Hold Gold If President Trump Is No More Irresponsible Than the Fed?

Enthusiasts cite other major risks which, in their view, justify the ownership of gold. Perhaps most importantly, President Trump has repeatedly expressed his desire to “stack” the Fed’s Board of Governors, and thereby to slash its policy rate of interest to a level which (in these critics’ view) triggers financial instability – or worse. John Cochran, a senior fellow of the Hoover Institution at Stanford University and an adjunct scholar at the Cato Institute, rebuts this and other criticisms.

“The policy world is aghast,” he writes (see “Trump’s Monetary Policy Desires Aren’t Crazy,” The Wall Street Journal, 21 October), “but Trump’s desires for monetary affairs aren’t as crazy as conventional wisdom portrays.” 

Cochrane sees three broad desires (for Trump and his supporters) and resultant opportunities (for owners of gold): “interest rates should be lower, in part to lower interest costs on the debt. The Federal Reserve should be less independent, subject to more democratic accountability. And ‘exorbitant privilege’ or ‘reserve currency status’ – that the world wants to hold our money and buy our debt, sending us goods in return – damages the U.S.”

The standard response from President Trump’s critics is that his desire for lower rates (these critics exempt Jerome Powell, who clearly also wants lower rates, and indeed cut the Federal funds rate on 29 October, from their disparagement) will trigger higher inflation. But, Cochrane asks, how soon and how much? “The best empirical estimates find that lower interest rates lead to no or slightly lower inflation for a year or so, then slightly higher inflation after two or three years. Even that response is barely significant statistically. And ... (the best estimates) say little about persistently lower interest rates.”

He astutely adds: “mainstream, or ‘new Keynesian,’ economic theory predicts that a permanently lower interest rate will eventually lower inflation, other things (especially fiscal policy) held constant, even though inflation may temporarily rise. This is an unsettling implication that the theory’s largely centre-left practitioners have trouble applying to reality, but there it is. Sure, maybe decades of consensus theory is all wrong. Economists have pursued wrong theories before. But if it’s in the equations of your own models, the proposition at least bears consideration.”

What about the proposition that the Federal Reserve, European Central Bank, RBA, etc., are “independent” and should remain so? In my view, it’s absurd (for details, see Why the RBA should be abolished – and what could replace it, 10 April 2023; see also Stop obsessing about the RBA, 14 February and Does the RBA’s “negative equity” matter? 20 April 2023).

Never mind what occurred before the COVID-19 pandemic, especially during the GFC, and focus on what’s happened since 2020: central banks have vastly expanded their scope of operations (and balance sheets), supported asset prices, channelled and funnelled credit, strayed into climate policy and inequality – and above all, monetised trillions of dollars of government debt. In these ways and more, they’ve slavishly followed governments’ wishes. So much for their “independence”!

President Trump is, in effect, demanding that the Fed continue to act irresponsibly. But he’s not insisting that it act more recklessly than it already has – under the guise of “independence”! – since the GFC and especially during and since COVID.

Will the results, if he gets his way, be appreciably worse than the Fed’s and other central banks’ so-called “independent” actions since the GFC?

Conclusions

Since August 1971 

1. gold has generally been relatively risky (that is, high-volatility) and low-return;
2. like individual stocks, equity market indexes and other commodities, gold’s short-term (12-month) fluctuations have been random and thus unpredictable;
3. its most recent short-term, medium-term (five-years) and long-term (ten years and beyond) returns are grossly unrepresentative of all short, medium and long-term returns over the past 54 years.

Over the past few years gold has certainly boomed; has it become a speculative mania? I’ve reviewed four sets of indicators. No definitive conclusion is possible, but it seems to me that the answer is more likely to be “yes” than “no.”

The old financial adage is probably right: “you never know you’re in a bubble until it bursts.”

Finally, analysing data compiled by the Cleveland branch of the U.S. Federal Reserve, I’ve undermined a key premise which has historically underpinned the desire to own gold: the expectation that consumer price inflation will accelerate to a high level. These data, which since the 1980s have reasonably accurately predicted CPI’s short- and medium-term rate of change, presently foresee modest – and decelerating – inflation.

That’s a big problem for bulls: buying gold makes no sense under these conditions. Moreover, buying it at current prices – and under the overconfident expectation that the continued stellar rise of its price is a “sure thing” – is aggressively chasing speculative returns. That’s a fool’s errand which typically ends in tears.

Neither I nor anybody else can credibly conclude – particularly if it’s become a speculative mania – that gold’s price can’t continue to rise. However, given that its 12-month returns have for decades been largely random and thus mostly unpredictable, it’s reasonable to infer that its return over the next 12 months will be much closer to its mean since August 1971 (CPI-adjusted 6.3%) than in the year to September 2025 (CPI-adjusted 47.4%).

Unless it’s different this time, losses, including severe losses, are also possible – and perhaps likely – over the next five years and beyond. Equally, neither I nor anybody else can credibly say that a crash is either nigh or inevitable.

Professionals – but few non-professionals – can, of course, mitigate gold’s risk of loss by hedging its price. Doing so, however, acknowledges my most fundamental point: holding gold no longer hedges other risks; instead, owning it has become a risk which requires hedging. Moreover, managing this risk can be expensive – particularly if you’re managing the risk of a falling price when the market is exuberant.

I conclude that, on balance, buying gold at its present price incurs a significant risk that you’ll strand or shrink your capital for years to come.

Implications for Conservative, Contrarian, Long-Term Investors

Leithner & Company doesn’t, figuratively speaking, join crowds or stand in queues. Quite the contrary: we take care to steer well clear of them. Booms, manias and the like are for speculators – and the vast majority of speculators eventually lose.

Instead, we shop where the crowds don’t; specifically, we seek out-of-fashion but quality assets selling at a discount. Just as fashions change, markets are cyclical: tomorrow’s returns are typically located where today’s aren’t – and vice versa.

Accordingly, we seek investments which lack queues – that is, whose sellers outnumber their buyers, and thus depress prices and returns.

Where should conservative, contrarian and long-term investors carefully search while the crowd enthusiastically chases yesterday’s news? Leithner & Co focuses upon established, well-financed companies in essential industries and with long histories of profitability who face short-term and even medium-term vicissitudes – and, as a consequence, unwarranted pessimism.

We’re content to leave speculators to their follies, and can afford to wait: thanks not least to speculators’ overall inexperience and overconfidence, financial markets eventually punish the aggressive and impulsive and reward the prudent and patient.

Click LIKE so that Livewire knows that you want more of this type of content. Click FOLLOW for notification when my next wire appears.