When Rates Rise, What Happens to Gold, Stocks, and Bonds

▪ WATCH THE VIDEOS, THEN READ ON

This written commentary goes with our two videos. Part One looks at gold, and Part Two looks at stocks and bonds. Watch them first, then read on for the full data behind every chart.

▶ PART ONE

When Rates Rise, What Happens to Gold?

▶ PART TWO

When Rates Rise, What Happens to Stocks and Bonds?

WATCHING ON YOUR PHONE? Tap the video, then tap the full-screen button (the square corners at the bottom right of the player). Turn your phone sideways for the biggest picture.

On September 16, the Federal Reserve raised its target range for the federal funds rate to 3.75 to 4.00 percent [1]. Whenever rates go up, the first question we hear from clients is some version of the same one: what does this mean for gold, and what does it mean for my stocks and bonds?

The written version follows the same data as the videos, with a little more room to explain it. After Part Two, we bring both together and explain how the KFSC Risk Managed Strategies are positioned today, and why.

All of the figures are as of September 26, 2026. Because that date fell on a Saturday, market prices are the close on Friday, September 25. Economic data is the latest released by September 26, and where a series ends earlier, we say so.

We are not trying to produce a trading signal. We want you to see what we see: the history, the current data, what supports our thinking, what challenges it, and what would actually change it.

Where gold stands today

Before we go into the history, here is the starting point.

Gold closed at $4,286.25 on September 25 [2]. That is 14.3 percent above where it closed twelve months earlier, 8 percent above its July 16 low of $3,969.94, and 20.6 percent below its January 28 high of $5,399.29. Measured from the December 31, 2025 close of $4,314.12, it is down 0.6 percent for the year.

All four figures use LSEG daily closing prices for gold spot (XAU=) [2]. These are changes in the price of gold, not the return of any account.

Those numbers can sound like they contradict each other. They do not. Gold can be well above where it was a year ago and still well below a recent high. The only thing that changes is where you start measuring, and that idea runs through everything that follows.

It also explains why two clients in the same strategy can look at their statements and feel very differently. Someone who started before the big rise and someone who started near the January high own the same strategy, with the same positioning. The investment did not change. The starting point did.

To get a clearer view, we look at the longer record.

What the same dollar did in gold, in silver, and in the S&P 500.

Suppose $1,000,000 had gone into each of gold, silver, and the S&P 500 on January 3, 2000, and then been left alone. By September 25, 2026, the gold would have grown to about 14.9 Million Dollars. The silver would have grown to about 12.4 Million Dollars, measured to September 18, the last date in our silver data. The S&P 500, with dividends reinvested, meaning the cash companies pay to shareholders was put back in, would have grown to about 8.7 Million Dollars[2][3].

Value of $1,000,000 placed in gold, silver, and the S&P 500 with dividends reinvested on January 3, 2000. Gold and the S&P 500 to September 25, 2026; silver to September 18, 2026 [2][3].

Two things are worth keeping in mind. January 2000 was close to the peak of the dot-com market, which makes it a hard starting point for stocks. And these are the metals and the index by themselves, before any cost, fee, spread, storage or tax. They are not the result of any KFSC account or strategy.

The same stretch also saw very large growth in money and government debt. M2, which is the money people and businesses hold in cash, checking, savings and similar accounts, grew from about 4.7 Trillion Dollars in January 2000 to about 23.3 Trillion Dollars in August 2026 [4]. Federal debt grew from about 5.8 Trillion Dollars to 39.5 Trillion Dollars in the second quarter of 2026 [5].

We pay attention to both because, over time, they affect what a dollar can buy. But neither one explains the price of gold by itself. Money grew at almost the same yearly pace through much of the 1980s and 1990s, and gold fell through much of that time.

Now shorten the window, and the answer changes.

What $1,000,000 became from four different starting dates. Gold and the S&P 500 with dividends reinvested to September 25, 2026; silver to September 18, 2026 [2][3].

So far in 2026, $1,000,000 became about $990,000 in gold, about $930,000 in silver, and about 1.14 Million Dollars in the S&P 500. Over the last twelve months, it became about 1.14 Million Dollars in gold, 1.58 Million Dollars in silver, and 1.19 Million Dollars in the S&P 500. From the end of 2019, before COVID, it became about 2.83 Million Dollars in gold, 3.71 Million Dollars in silver, and 2.65 Million Dollars in the S&P 500 [2][3].

Which of those numbers is the true one? All of them. The answer changes because the starting date changes. That is why we show several windows instead of choosing the one that makes the story look best, and why a long-term strategy should not be rebuilt around whichever period happened to look best, or worst, most recently.

What the record shows

Twelve completed periods of rising rates, and what gold did in each.

Since 1971, we identified twelve completed periods in which the federal funds rate, the short-term interest rate the Federal Reserve controls, rose by at least 1.5 percentage points, and we had a gold price at both ends. Gold finished higher in nine of the twelve. The middle result was a gain of 9.3 percent[2][6].

That surprises people who assume rising rates automatically mean falling gold. The two largest gains, 114.5 percent and 278.1 percent, came during two of the fastest increases in the federal funds rate, from 3.3 to 10.78 percent and from 4.61 to 17.61 percent.

Change in the gold price across each completed period of rising rates since 1971 [2][6]. We show every period rather than a selected few.

That does not mean rising rates are good for gold, and it does not tell us what happens next. What it tells us is narrower: the direction of interest rates, by itself, has not been enough to explain the direction of gold. The reason rates were rising, how fast prices were rising at the time, and what cash earned after inflation all mattered.

One limit matters. We can only identify the end of a rate cycle after it has happened, so this record helps us understand history. It is not something anyone could have used at the time to trade.

You will see two counts, twelve and thirteen. Twelve periods have a gold price at both the start and the end. A thirteenth ran from July 1967 to May 1968, but freely traded gold prices begin only near the end of it. So it has no start-to-finish result, but there is enough price history in its final weeks to measure a decline inside it. That is why there are twelve results, but thirteen periods when we look at declines.

There is another way to look at the same question. Instead of choosing completed rate cycles, we took every 24-month stretch in the record. Think of it as sliding a two-year window across the whole history, one month at a time, and sorting each window by what the federal funds rate did. When the rate rose by more than one percentage point, the middle change in gold was about +18 percent, and gold was higher 80 percent of the time. When the rate barely moved, the middle change was −3.2 percent, higher 44 percent of the time. When the rate fell by more than one point, it was +17.1 percent, higher 63 percent of the time [2][6]. These windows overlap, so they are far fewer than hundreds of separate tests, and they do not point to any rate level at which gold must do something next.

The period gold holders cannot ignore

The strongest case against gold, stated directly.

Three of the twelve periods ended with gold lower, and all three fell between 1980 and 1989. Any fair reading of the record has to spend time there.

First, what came before it. Until August 1971, the official price of gold was $35 an ounce [7]. That changed when President Nixon ended the link between the dollar and gold. By January 21, 1980, gold had closed at $835, an increase of about 2,286 percent in less than nine years [2]. So the decline that followed started from an extraordinary peak.

From that close, gold fell to $252.30 on August 25, 1999, a decline of 69.8 percent over 235 months [2]. Measured in what the money could actually buy after the rise in consumer prices, the decline was about 85.9 percent[2][8]. Gold did not get back to its January 1980 price until December 28, 2007, and after inflation it took until September 2025. Even at the 1999 low, gold was still about seven times higher than the $35 it was fixed at before 1971. And stocks have had deep falls too: from 1929 to 1932, stock prices fell about 85 percent [9].

What the record shows from January 1980 to August 1999 [6][10][11][8][4][5][2][7].

Several things were working at the same time. Cash paid far more than inflation. In June 1981, the federal funds rate reached 19.1 percent while consumer prices were rising about 9.7 percent from a year earlier, a gap of about 9.4 percentage points[6][10]. Put simply, you could keep money somewhere safe and earn far more than prices were rising. Across 41 months in the 1980s, the federal funds rate was more than five points above inflation.

It also lasted. For almost twenty years, the 2-year Treasury paid more than inflation in 96 percent of months, by about 3.7 points on average [11][8]. Inflation came down, from almost 15 percent in the year to April 1980 to about 3 percent a year in the 1990s. And gold started from a peak.

Gold pays no interest. Think of two choices side by side. One pays you interest every month. The other pays nothing, and only helps you if its price rises. When the one that pays is paying well above inflation, year after year, gold has a much harder time competing. That is why we do not dismiss interest rates.

What today’s rate actually means

The rate alone does not tell you what a saver is gaining.

A 4 percent interest rate does not tell us how much buying power a saver is gaining. We also need to know how fast prices are rising. Economists call the difference the real rate, and it is the number we watch, not the headline rate.

In August 2026, the effective federal funds rate was 3.63 percent, and consumer prices were 3.4 percent higher than a year earlier [6][10]. After the September 16 increase, the middle of the new range, 3.875 percent, is about half a point above that inflation reading [1][10]. The 2-year Treasury, at 4.81 percent on September 25, pays about 1.5 points more than inflation [11][10].

Compare that with June 1981, when the federal funds rate was 9.4 points above inflation, or with the twenty years from 1980, when the 2-year Treasury averaged about 3.7 points above it. Since 1999, the 2-year Treasury has paid no more than inflation in about 53 percent of months. From 1977 to 1998, that happened in only about 10 percent of months [11][8].

So today’s relationship is much closer to the last twenty-five years than to the 1980s. That is a reading of the present, and it can change.

A large decline does not automatically change the long-term role.

How does the 2026 fall compare with the ones before it?

Gold’s decline in 2026 has been significant. From its January 28 high of $5,399.29, it fell 26.5 percent to its July 16 low of $3,969.94. By September 25, it had recovered 8 percent from that low, but it was still 20.6 percent below the January high [2].

Put that next to the longer climb. Gold reached a low of $1,621.57 on September 26, 2022, and from there rose about 233 percent over forty months to the January high [2]. The recent decline matters. So does the longer climb. Both are part of the story.

Against the longer record, falls of this size have been common while rates were rising. Eleven of the thirteen rate-rise periods had a fall in gold of 10 percent or more somewhere inside them, and the middle fall was about 21 percent[2][6]. A period can end with gold higher and still hold a very uncomfortable decline along the way.

The five major high-to-low declines in gold since 1970, from the highest close to the lowest [2].

We identified five major declines in gold, from a high to a low, since 1970. From 1974 to 1976, gold fell 47.7 percent over twenty months. From 1980 to 1985, 66 percent over sixty-one months. From 1996 to 1999, 39.2 percent over forty-two months. From 2011 to 2015, 44.6 percent over fifty-one months. And from January to July 2026, 26.5 percent [2].

So far, 2026 is the smallest of the five. But notice the words so far. The first four are complete, and we know with hindsight where their lows were. The 2026 decline is still unfolding. We do not know whether July was the final low. History gives us perspective on the size and length of earlier declines. It does not give us a prediction.

Three periods worth knowing

Why the first rate increase does not settle what comes next.

1974. Gold fell from $179.80 on April 3 to $134.30 on July 5, a drop of 25.3 percent, while the federal funds rate climbed to its 12.92 percent high in the same month gold hit its low. Twelve months later, gold was 22.2 percent higher. The deeper decline came afterward, from $195.50 on December 30, 1974, to $102.20 on August 30, 1976, down 47.7 percent, while the federal funds rate fell from 8.53 to 5.29 percent [2][6]. The first decline happened while rates were rising. The deeper one happened while they were falling.

2011 to 2015. Gold fell from $1,898.99 on September 5, 2011, to $1,051.36 on December 17, 2015, a decline of 44.6 percent over fifty-one months. The federal funds rate started at just 0.08 percent and ended at 0.24 percent [2][6]. The Federal Reserve did not raise rates until December 16, 2015, essentially at the very end, when gold closed at $1,072.56. The largest single leg of the fall, 32.9 percent, ran from October 4, 2012, to June 27, 2013, while the federal funds rate stayed between 0.09 and 0.16 percent. Gold lost almost half of its value from that high while short-term rates sat near zero.

2022. On March 16, 2022, the Federal Reserve made its first increase of the cycle, with gold at $1,927.93. Gold fell a further 15.9 percent to $1,621.57 on September 26. Then it recovered: up 17.2 percent over the next twelve months and 64.7 percent over the next twenty-four, while the Federal Reserve kept raising rates to about 5.33 percent in August 2023 [2][6]. Gold reached its low roughly ten months before the final increase.

Put together, gold has fallen while rates were rising, fallen hard while rates were near zero or falling, and started recovering while the Federal Reserve was still raising rates. There is no dependable sequence. And a low is obvious only after the fact. Nobody living through these declines got a sign announcing the bottom.

The closing price at each major low and the closing price 36 months later. The 2026 line covers two months and is shown for scale only [2].

Once we look back and name those lows, what happened over the next three years? After the August 1976 low of $102.20, gold was $319.40 three years later, up about 212.5 percent. After the February 1985 low of $284.20, about $433, up 52.3 percent. After the August 1999 low of $252.30, up about 21.6 percent. After the December 2015 low of $1,051.36, up about 18.5 percent [2]. Very different outcomes, and no single recovery pattern.

The 2026 line is there only so you can see where we stand. Gold is 8 percent above its July low after about two months, not thirty-six. We are not pretending two months compares with three completed years, and none of this suggests 2026 must follow any earlier period.

Which interest rate?

The Federal Reserve controls one rate. The market prices the others.

When people hear that rates are going up, they usually picture one interest rate. There are several, and they do not always move together.

The federal funds rate is the one rate the Federal Reserve actually sets. It is the rate banks charge each other to borrow overnight. On September 16, the Federal Reserve raised its target range to 3.75 to 4.00 percent, from 3.50 to 3.75 percent [1].

Treasury yields are different. A Treasury is simply a loan to the U.S. government, and the yield is the interest it pays. On September 25, the 2-year Treasury yielded 4.81 percent and the 10-year yielded 5.17 percent[11][12]. The Federal Reserve does not set those. Investors do, every day, by deciding how much interest they want before they will lend to the government for two years, or for ten.

A good example: on September 25, both yields went down, from 4.87 and 5.18 percent the day before [11][12]. The Federal Reserve did nothing that day. And it is the longer rates, like the 10-year, that have the most to do with what you pay on a mortgage, what a business pays on a loan, and what investors are willing to pay for stocks. So when someone says rates are rising, the first question to ask is, which rate?

Why older bonds lose value when rates rise

Same bond, same payments, different price.

A bond is a loan. You lend money, and in return you get a fixed interest payment, called the coupon, and your money back at the end.

Picture two 10-year Treasury notes. Same borrower, same ten years. The older one pays 3 percent. A new one pays 5 percent. If you own the older note and want to sell it, who will pay you full price for 3 percent when they can buy 5? Nobody. So the price of the older note falls until a buyer would earn about the same on either one.

Think of it like a car. If the new model gets much better gas mileage for the same price, the older model is worth less. Nothing happened to the old car. What changed is what you can get today. That is why a bond can show a lower value on your statement even though it pays the same interest it always did. There is a good side, too: when rates are higher, the next bond you buy pays you more.

How long a bond has left to run makes a big difference. Take three Treasuries, with 2, 10 and 30 years left, all paying 3 percent, and then let market rates rise to 5 percent. The 2-year is stuck with the lower rate for only two more years, so its price falls the least. The 30-year is stuck for thirty years, so its price falls the most. It is like a landlord: a 2-year lease signed at the old rent does not hurt much when rents rise; a 30-year lease at the old rent hurts a lot more. If rates fall, it works the other way, and the 30-year gains the most.

The notes and Treasuries in this section are hypothetical examples to show the arithmetic. They are not quotations for any security and not a forecast for rates.

Higher rates change what money later is worth today

The same arithmetic sits under bonds, houses, and stocks.

Here is the idea in one question. If you want $100,000 ten years from now, how much do you have to put into a 10-year Treasury today?

Each amount is $100,000 divided by (1 + yield) once for each of ten years, with every payment reinvested at the same yield. Yields are the 10-year Treasury on the dates shown [12]. Illustrative arithmetic, not a quotation and not a forecast.

The higher the rate, the less you need to set aside today for the same amount later. That matters for stocks, because when you buy a stock you are really buying a share of a company’s profits for years to come. That is money that shows up later. When a safe Treasury pays more, investors are willing to pay less today for money that shows up later.

A rental house makes it easy to see. Say you have $100,000. You can buy a 10-year Treasury, or a small rental house that brings in $5,000 a year. A year ago, the Treasury paid about $4,180 a year, so the house paid more and a $100,000 price made sense. Today the Treasury pays about $5,170 a year, backed by the U.S. government, and the rent is still $5,000 [12]. At today’s rate, that same $5,000 of rent is worth about $96,712 ($5,000 divided by 5.17 percent). A company’s profits are like the rent. When Treasuries pay more, investors tend to want to pay less for the same profits.

The house is a made-up example. Real property prices depend on much more than interest rates, including costs, vacancies, taxes, and changes in rent.

Why the Treasury market matters beyond bonds

The government must borrow, and the price lenders demand reaches the whole economy.

Federal debt is about 39.5 Trillion Dollars at the end of June 2026 [5], or about 122.6 percent of everything the economy produced in a year as of the end of March [13]. Over the twelve months to August 2026, the government spent about 1.8 Trillion Dollars more than it collected [14].

It covers that gap, and pays off old Treasuries as they come due, by selling new ones. Somebody has to buy all of them, and the buyers decide how much interest they need before they will. Money market funds are one of the largest buyers of short-term government debt. They held about 8.44 Trillion Dollars at the end of June [15]. That money belongs to their investors, and it can move. Money market funds are not insured by the FDIC, are not guaranteed, and can lose value.

If the government has to pay more to find enough buyers, it does not stay in the bond market. Treasury yields are the starting point for much other borrowing, so higher yields can mean higher mortgage rates, more expensive business loans, tighter credit, and pressure on what investors are willing to pay for stocks. That is why we watch this market so closely.

Four ways higher rates reach stocks

Rates add pressure. They do not decide the outcome on their own.

These do not all hit at once or with the same force. And they are not the only things that matter. Company profits, how fast the economy grows, inflation, and how expensive stocks were to begin with all shape the result.

What stocks did when the Fed raised rates

The same twelve periods we used for gold, measured the same way.

In nine of the twelve rising-rate periods since 1971, the S&P 500 finished higher. The middle result was a gain of about 10.2 percent[6][16].

But in every one of the twelve, stocks fell somewhere along the way. The smallest drop was 7.6 percent, the middle drop about 12.9 percent, and the largest 25.4 percent. That largest drop came in the most recent period, from December 2021 to August 2023, and the index still finished that period up 1.4 percent [6][16]. For a client, both parts matter: where it ended, and what you had to live through on the way.

S&P 500 price change and largest decline in each of the twelve rising-rate periods since 1971. Price only, dividends not included [6][16].

The periods also lasted very different lengths of time, from five months to sixty-three months, with the middle one about a year and a half. Length did not sort the results. The longest period, 2014 to 2019, had the largest gain, 55.1 percent. The second longest, 1977 to 1980, ended down 4.9 percent [6][16]. The Federal Reserve raised rates again in September. This history cannot tell us whether this turns into a short period or a long one, which is why we do not build the strategies around a guess about timing.

Inflation made a big difference

When inflation stayed lower, stocks tended to finish higher.

When inflation averaged under 6 percent during a rising-rate period, the S&P 500 finished higher in seven of eight periods, with a middle gain of about 14.4 percent. When inflation averaged 6 percent or more, it was two of four, with a middle result of about −1.8 percent[6][16][10].

Average inflation and the S&P 500 price change in each of the twelve rising-rate periods [6][16][10]. The 6 percent line only groups the periods.

There is one big exception. From 1980 to 1981, inflation averaged about 11.7 percent and stocks still rose 15.2 percent. We also want to be clear about the limits. High rates and high inflation tended to arrive together, so this chart cannot separate one from the other, and four high-inflation periods is a very small sample. There is nothing special about 6 percent; it simply splits the twelve into two groups.

Today, inflation is running at about 3.4 percent a year [10]. That is well under 6 percent, but it has not gone away, so we watch it closely.

Valuations are high. They are not a timing tool.

The price you start from matters.

The Shiller CAPE ratio compares the price of the stock market with what companies earned over the previous ten years, after inflation. Think of it as the price tag on each dollar of earnings. In September 2026, it stood at 40.58. Only 19 of the 1,869 months since 1871 were higher, which puts today above about 99 percent of the record. The middle reading over that whole time is about 16.1. Today is also above the 1929 peak of 32.56 [17].

The S&P 500 Shiller CAPE ratio, monthly, January 1871 to September 2026. Before March 1957 the series uses the indexes that came before the S&P 500 [17][9].

Here is how we think about it. If you pay a lot more than similar houses usually sell for, it does not mean the price falls tomorrow. But you have a lot less room for error if something goes wrong. That is what a high starting price means for stocks.

Two things are true at the same time, and holding only one of them is where investors get hurt. The first is that fundamentals should not be ignored. After the December 1999 reading of 44.20, the ratio fell to 13.32 by March 2009, a decline of about 70 percent in the ratio itself [17]. A ratio can fall because prices fall, because earnings rise, or both, so that is not a client return. It shows how far the price paid for each dollar of earnings can move.

The second is that a high reading is not a reason to act on its own, because a market can stay expensive for a very long time. The ratio has been above its 1929 peak for 32 straight months, since February 2024. It has been above its 1966 peak of 24.06 for 127 straight months, since March 2016. In the late 1990s, it sat above 30 for 51 straight months, from June 1997 to August 2001 [17]. Someone who treated the June 1997 reading as a signal would have watched it rise for more than four years before it turned.

Only two of the twelve rate periods started anywhere near today’s level, in January 1999 at 40.58 and December 2021 at 38.3. Stocks finished higher in both, after falling 12.1 percent and 25.4 percent along the way. Over the longer record since 1971, starting in the most expensive quarter of months was followed by about 4.9 percent a year over the next ten years, price only; starting in the cheapest quarter, by about 11 percent a year [17][16]. Those ten-year stretches overlap, so they amount to about five separate decades, not hundreds of separate tests.

A second measure tells a similar story. Take the value of all U.S. stocks and divide it by the size of the U.S. economy, often called the Buffett Indicator. Since 1970, the middle reading has been about 90 percent, a little less than one year of everything the economy produces. On September 18, it was about 234 percent, the highest reading since 1970 [18].

Total U.S. stock market value divided by gross domestic product, daily, December 1970 to September 18, 2026 [18].

High readings can last. The measure has been above its 2000 peak of 149 percent since October 2023, and above its 2007 peak of 109 percent since March 2020 [18]. Neither measure tells us when markets will rise or fall. Together they tell us that stocks are starting from an expensive place, and that is one reason we are careful about how much stock market risk we carry.

Stocks since the September rate increase

Nine days is not a rate cycle.

The day before the increase, September 15, the S&P 500 closed at 7,585.73. It dipped on the day of the increase, climbed for a few days, and gave a little back. By September 25 it closed at 7,743.41, about 2.1 percent above the close before the increase, about 0.3 percent below its highest close since the increase (7,764.70 on September 21), and about 0.7 percent below its highest close of 2026, 7,798.99 on August 13 [1][19]. Those are closing prices, price only. Nine days tells us where we are starting from, not what comes next.

There is no risk-free asset

It is not just what you hold. It is when you hold it, and knowing why.

By buying power, we mean what your money can actually buy after inflation. Every major way to hold money has lost a large share of it at some point.

Largest loss of buying power for stocks, gold, 10-year Treasuries and the dollar, measured with the purchasing power of the dollar [9][2][12][8]. Treasury figures are estimates with interest reinvested.

Stocks. From 1929 to 1932, stock prices fell about 85 percent, and about 81 percent after inflation. Stocks had been bought at about twice their long-run valuation, and then earnings and the economy collapsed. It took until the 1950s to recover [9][8].

Gold. From 1980 to 1999, gold lost about 86 percent of its buying power, and did not get all of it back until September 2025 [2][8].

10-year Treasuries. From March 1967 to September 1981, even with every interest payment reinvested, they lost about 47 percent of their buying power, because inflation ran ahead of the interest. They recovered in October 1985 [12][8].

The dollar. A dollar from 1913 buys about 3 cents’ worth today, a loss of about 97 percent of its buying power, and since August 1971 it has lost about 87.8 percent [8]. Interest on savings makes up part of that, but only part.

The starting date matters here too. If $100,000 went in at the start of 2000, gold now buys almost 7.8 times what it did, though someone who bought at the 2011 high waited until October 2024 to get back to even. Stocks with dividends reinvested buy about 4.3 times as much, after losing 54 percent of their buying power from 2000 to 2009 and recovering in May 2013. 10-year Treasuries, with interest reinvested, are about 1.35 times. Cash earning nothing buys about half of what it did [2][3][19][12][8].

Start at the beginning of 2020 instead, and the order changes. Gold buys about 2.25 times what it did, after falling 31.2 percent along the way. Stocks with dividends reinvested about doubled their buying power, after a fall of 28.5 percent from 2021 to 2022. Treasuries bought in 2020, when the 10-year paid barely half a percent, now buy only about 0.74 times what they did and have not recovered, while cash lost almost a quarter of its buying power [2][3][12][8]. Even something backed by the U.S. government can lose buying power when it is bought at a very low rate.

Buying-power values use month-end figures through August 2026. Treasury results are estimates: a Treasury bought each month at the month-end yield, repriced the next month, with interest reinvested. They are not an index or a traded price, and ignore costs and taxes.

Why we hold more gold and less in stocks right now

We run both through the same four frameworks.

Many of you have asked us this directly. We answer it the same way we answer every portfolio question: we run the data through the same four KFSC frameworks. Think of them as four questions we keep asking about the world.

In our current reading, all four point toward gold more clearly than toward stocks. That is why we hold more gold and less equity exposure. It does not mean gold cannot fall or that stocks cannot rise. The frameworks help us read the environment. Our advisors decide the positioning.

It also helps to see why we do not treat gold and stocks the same way. When you own stocks, you own a piece of real businesses and share in their profits. That is where the return comes from, and where the risk comes from too. So when we want to take less risk, stocks are usually the first place we adjust. Gold does not depend on any company’s profits. We hold it because of what is happening to money itself. It responds to interest rates after inflation, the dollar, government debt, and confidence in the financial system.

In Part One, gold finished higher in nine of the twelve rising-rate periods, the same count as stocks, but its falls along the way were usually deeper. Gold pays no interest, it can fall hard, and it can fall at the same time as stocks. So the two do different jobs and are managed differently. Stocks are where we usually turn risk up or down. Gold is judged against the reasons we hold it, not only against its price.

Several pressures at the same time

Each has happened before. What concerns us is that all four are here together.

Debt. Federal debt is about 122.6 percent of everything the economy produces in a year. In 1999, it was about 60 percent [13]. Deficits. The government is spending at a rate of about 7.76 Trillion Dollars a year and collecting about 6.06 Trillion Dollars [20][21]. Think of a household that spends $128 for every $100 it brings home. Rates. The 10-year Treasury yield is near its highest close since July 2007, and the Federal Reserve has just raised rates again [12][1]. As old debt is replaced at higher rates, the interest bill goes up. Valuations. Stocks are near the most expensive levels in their history on both measures we looked at [17][18].

These can feed on each other. None of them tells us that stocks must fall. Together, they tell us the risk is higher, and that is why we are positioned defensively in stocks.

Why gold remains a structural holding

Its volatility is easy to see. The conditions it is held for move more slowly.

What has happened to the dollar. A dollar today buys about 12 cents’ worth of what it bought in 1971. Over that time, after inflation, gold grew about 13.2 times and the S&P 500 about 9.5 times, on prices alone [8][2][19]. With dividends, the stock figure would be higher, and we want to be fair about that.

Cash and inflation. Since 1999, the 2-year Treasury has paid no more than inflation in about 53 percent of months, compared with about 10 percent from 1977 to 1998 [11][8]. For most of the last twenty-five years, holding cash often did not protect what money could buy.

Central banks. They are still adding gold. China’s central bank has bought gold for twenty-one straight months to July 2026: about 20 tonnes in July and about 60 tonnes so far in 2026, bringing its gold to about 2,366 tonnes, 8 percent of its reserves. Central banks together bought about 130 tonnes in 2026 through July, slower than the roughly 160 tonnes in the same months of 2025 [22].

Reserve holders. China’s recorded holdings of U.S. Treasuries fell about 52 percent from January 2014 to July 2026, from about 1.28 Trillion Dollars to about 0.62 Trillion Dollars [23]. Part of that reflects lower bond prices, not only sales, and holdings through other countries may not be captured.

Purchasing power, cash after inflation, central bank buying, and China’s Treasury holdings [8][2][19][11][22][23].

None of this means gold cannot fall. It pays no interest, and when interest rates after inflation are high, that competes with it. But in our view, the reasons we hold it are still in place. How much we hold can change as the data changes, but not simply because the price moves.

Why we hold short-term Treasuries

A parking spot that pays you while you wait.

Short-term Treasuries let us stay invested in gold and stocks while helping to reduce how much the whole portfolio swings. They pay interest: the 2-year yielded 4.81 percent on September 25 [11]. They come due quickly and are easy to sell, so money is available if prices fall or the data changes. Your advisor decides when.

Their prices are steadier. If rates rise one point, the price of a newly issued 2-year Treasury falls about 1.9 percent, against about 7.4 percent for a 10-year. In the worst stretch, 2020 to 2023, the estimated 2-year fall was about 6.1 percent and the 10-year about 26.3 percent. The 2-year recovered by May 2024; the 10-year has not [11][12]. They are backed by the U.S. government, so default risk is very low, though not zero, and the 2-year now pays about 1.5 points more than inflation [11][10].

Short-term Treasuries: income, price stability, default risk, and a hypothetical illustration of portfolio swings since 2000 [11][12][10][3][2][8].

The example on the right of that slide is hypothetical and is not our allocation. Splitting money three ways, between gold, stocks and 2-year Treasuries, rebalanced monthly from January 2000, swung about 7.8 percent a year against 11.6 percent for half gold and half stocks, and its worst drop was about 16.8 percent against 25.3 percent. It also grew more slowly, about 7.7 percent a year against 10.2 percent [2][3][11]. It was built with hindsight and includes no fees, taxes or costs.

Short-term Treasuries have limits. If rates fall, maturing Treasuries are reinvested at lower yields. Inflation can rise above what they pay again, as it did in more than half of months since 1999. And money held in reserve can fall behind stocks if markets keep rising.

Scarcity beyond gold

A rising price does not prove a shortage.

Our Strategic Scarcity framework tracks the physical supply of the things the economy runs on. Silver is one example. For five years in a row, 2021 through 2025, more silver was used than was mined and recycled, about 647 million ounces in all. Mine output slipped from about 824.3 to 804.9 million ounces, while silver used in solar panels rose from about 111.6 to 193.8 million ounces. To be fair, the shortfall narrowed sharply in 2025, to about 52.5 million ounces from 175.5 million in 2024 [24]. In our strategies, silver is treated as a risk asset, not a structural holding like gold.

Copper shows the other side. Its price rose about 44 percent over the year to August 2026 [25], yet the latest data show more refined copper produced than used, a surplus of about 509,499 tonnes from January to July 2026 [26]. The price and the physical data can tell different stories, which is why we watch both. We also follow uranium, rare earths, platinum and palladium, and energy. These markets can swing sharply, and supply can respond. None of this is a forecast.

What would make us reconsider gold’s role

No single price move, meeting, or headline. A sustained change in the data.

We look at many factors that could lead us to increase or decrease our gold position, and how much of the portfolio it takes up. But one of them would matter on its own. If cash and Treasuries were paying well above inflation, and that looked sustainable, that alone would be a break in our thesis, and we would reposition. We would also pay close attention if central banks became steady sellers of gold, because that adds supply.

That is not where we are today. The 2-year Treasury pays about 1.5 points above inflation, against about 3.7 points on average from 1980 to 1999 [11][8]. Central banks have been buying gold, not selling it [22]. We watch this closely, and we will tell you if it changes.

We would not change gold’s structural role because of one Federal Reserve meeting, one period of volatility, or one decline from a market high. And a strategy reaching a drawdown-review point does not trigger an automatic change. It prompts Ernesto and Emmelis to review the whole strategy and decide whether any change is appropriate. Gold exposure differs across the six KFSC Risk Managed Strategies according to each strategy’s risk mandate, and positions are managed at the strategy level, not separately for each account.

Where we stand today

Our current strategy-level positioning, not a forecast.

The frameworks are reading heavy debt and large deficits, higher borrowing costs, tight physical supply in some markets, and expensive stocks. Taken together, that has us in a defensive stance.

This positioning can still lose money, and it can trail a rising market. Gold can fall too. The model does not trade or rebalance accounts. Our advisors review the frameworks, portfolio risk, and each strategy before deciding what, if anything, changes.

We are also watching ongoing geopolitical conflicts and the U.S. federal election on November 3, 2026. They may add volatility, and they may also bring opportunity, but they do not decide our positioning by themselves. We know the swings in metals and in markets may continue, and our positioning takes that into account.

What you should take from this

Five points that bring the written commentary and the videos to the same place.

When the data changes materially, we reassess. If that leads to a strategy-level change, we will explain what changed, why it matters, and what we are doing about it.

That is what we mean by Models diagnose. Advisors decide. Portfolios implement. The model and the research help us organize the data. The decision remains an advisor decision.

When your own goals, risk tolerance, time horizon, liquidity needs, or financial circumstances change, please call us. That is how we make sure the KFSC Risk Managed Strategy you selected continues to fit you.

Sources

This commentary was prepared on October 5, 2026, using data as of September 26, 2026. Market figures are the close on September 25, 2026; economic figures are the latest released by September 26, 2026, and each source below states its own dates. References are numbered in order of first appearance. Percentage, window and buying-power computations were performed by Keaney Financial Services Corp directly from the exported series listed here.

[1] Federal Reserve Bank of St. Louis, FRED. Federal Funds Target Range - Lower Limit (DFEDTARL) and Upper Limit (DFEDTARU), daily, not seasonally adjusted, December 2008 to September 25, 2026 [Data set]. Cited for: the target range of 3.75 to 4.00 percent set on September 16, 2026, the prior range of 3.50 to 3.75 percent, and the date of the increase.

[2] LSEG Workspace. Daily series: gold spot (XAU=), 22 March 1968 to September 18, 2026 (exported September 18, 2026), and 20 February 2026 to October 5, 2026 (exported October 5, 2026) [Data exports]. Cited for: every gold price level and every change in the gold price, using daily closes through September 25, 2026; the depth of each decline and the dates of each high and low.

[3] LSEG Workspace. Daily series: silver spot (XAG=), 1 July 1982 to September 18, 2026 (exported September 18, 2026), and S&P 500 total return, dividends reinvested (.SPX), 4 January 1988 to September 25, 2026 (exported October 2, 2026) [Data exports]. Cited for: every silver price level and every change in it (to September 18, 2026), and every change in the S&P 500 with dividends reinvested (to September 25, 2026). The S&P 500 series is a cumulative total-return series; two rows dated December 1986 and December 1987 carry a different base and were excluded.

[4] Federal Reserve Bank of St. Louis, FRED. M2 Money Stock (M2SL), monthly, seasonally adjusted, January 1959 to August 2026 [Data set]. Cited for: the level of the money supply at January 2000 and at August 2026.

[5] Federal Reserve Bank of St. Louis, FRED. Federal Debt: Total Public Debt (GFDEBTN), quarterly, not seasonally adjusted, first quarter of 1966 to second quarter of 2026 [Data set]. Cited for: the level of federal debt at the first quarter of 2000 and the second quarter of 2026.

[6] Federal Reserve Bank of St. Louis, FRED. Federal Funds Effective Rate (FEDFUNDS), monthly, not seasonally adjusted, 1954 to August 2026 [Data set]. Cited for: the level of the rate at every date, the twelve periods of rising rates and their start and end dates, and the rate compared with the rate of price increases.

[7] Gold Reserve Act of 1934 and Presidential Proclamation 2072 (January 31, 1934), which set the official U.S. gold price at $35 an ounce; the price held until the United States ended the dollar’s convertibility into gold on August 15, 1971. Cited for: the $35 official gold price before August 1971.

[8] Federal Reserve Bank of St. Louis, FRED. Purchasing Power of the Consumer Dollar in U.S. City Average (CUUR0000SA0R), monthly, not seasonally adjusted, January 1913 to August 2026 [Data set]. Cited for: every figure measured in what the money would buy, including gold from 1980 to 1999, the buying-power comparisons since 1913, 1971, 2000 and 2020, and inflation compared with the 2-year Treasury yield.

[9] Shiller, R. J. U.S. Stock Markets 1871-Present and CAPE Ratio (ie_data.xls), monthly, January 1871 onward [Data set]. Cited for: the stock market’s largest fall, September 1929 to June 1932 (monthly prices 31.3 to 4.77), and the linked index history before March 1957.

[10] Federal Reserve Bank of St. Louis, FRED. Consumer Price Index for All Urban Consumers: All Items in U.S. City Average (CPIAUCSL), monthly, seasonally adjusted, 1947 to August 2026, updated September 11, 2026 [Data set]. Cited for: the rate of price increases at every date, including 3.4 percent in the year to August 2026, and the federal funds rate compared with it.

[11] Federal Reserve Bank of St. Louis, FRED. Market Yield on U.S. Treasury Securities at 2-Year Constant Maturity (DGS2), daily, June 1976 to September 25, 2026 [Data set]. Cited for: the 2-year yield of 4.81 percent on September 25, 2026, the 2-year yield compared with inflation, and the estimated 2-year Treasury returns.

[12] Federal Reserve Bank of St. Louis, FRED. Market Yield on U.S. Treasury Securities at 10-Year Constant Maturity (DGS10), daily, January 1962 to September 25, 2026 [Data set]. Cited for: the 10-year yield of 5.17 percent on September 25, 2026, the yields on the dates in the present-value examples, and the estimated 10-year Treasury returns.

[13] Federal Reserve Bank of St. Louis, FRED. Federal Debt: Total Public Debt as Percent of Gross Domestic Product (GFDEGDQ188S), quarterly, seasonally adjusted, first quarter of 1966 to first quarter of 2026 [Data set]. Cited for: federal debt of 122.6 percent of gross domestic product in the first quarter of 2026, and the same measure in 1999 and 2020.

[14] Federal Reserve Bank of St. Louis, FRED. Federal Surplus or Deficit [-] (MTSDS133FMS), monthly, not seasonally adjusted, October 1980 to August 2026 [Data set]. Cited for: the federal deficit over the twelve months to August 2026.

[15] Federal Reserve Bank of St. Louis, FRED. Money Market Funds; Total Financial Assets, Level (MMMFFAQ027S), quarterly, not seasonally adjusted, fourth quarter of 1945 to second quarter of 2026 [Data set]. Cited for: money market fund assets at the end of June 2026.

[16] LSEG Workspace. S&P 500 Index (.SPX) price history, daily close, 4 January 1971 to 30 January 2026 [Data export]. Cited for: the S&P 500 price change and largest decline in each of the twelve rising-rate periods, price only.

[17] GuruFocus. (2026, October 2). S&P 500 Shiller CAPE Ratio, monthly, January 1871 to September 2026 [Data export]. September 2026 reading 40.58; an earlier export (September 21) showed 40.68 before revision. Cited for: the level of the ratio in September 2026, its rank among the months of the record, its median, its earlier peaks, and the starting values in the ten-year comparison.

[18] GuruFocus. (2026, September 21). Buffett Indicator: total market value to gross domestic product, daily, December 1970 to September 18, 2026 [Data export]. Cited for: the level on September 18, 2026, the median since 1970, the readings at earlier peaks, and the dates from which the measure has stood above its 2000 and 2007 levels.

[19] LSEG Workspace. (2026, October 2). S&P 500 Index (.SPX) daily close, price only, 5 April 1934 to September 25, 2026 [Data export]. Cited for: S&P 500 closing levels in September 2026, and every price-only S&P 500 comparison since 1971 and since 2000.

[20] Federal Reserve Bank of St. Louis, FRED. Federal Government: Current Expenditures (FGEXPND), quarterly, seasonally adjusted annual rate, first quarter of 1947 to second quarter of 2026 [Data set]. Cited for: federal spending at an annual rate of about 7.76 Trillion Dollars in the second quarter of 2026.

[21] Federal Reserve Bank of St. Louis, FRED. Federal Government Current Receipts (FGRECPT), quarterly, seasonally adjusted annual rate, first quarter of 1947 to second quarter of 2026 [Data set]. Cited for: federal receipts at an annual rate of about 6.06 Trillion Dollars in the second quarter of 2026.

[22] World Gold Council. (2026, September). Central Bank Gold Statistics: Central banks make positive headlines on gold. Data from the IMF and respective central banks, to 31 July 2026 [Report]. Cited for: central bank gold purchases: about 130 tonnes in 2026 to the end of July, about 160 tonnes in the same months of 2025, and China’s purchases and reserves.

[23] Federal Reserve Bank of St. Louis, FRED. Foreign Portfolio Holdings of U.S. Long-Term and Short-Term Treasury Securities: China, Mainland (FORTREASPOS41408), monthly, not seasonally adjusted, February 2003 to July 2026 [Data set]. Cited for: China’s recorded holdings of U.S. Treasuries, January 2014 to July 2026.

[24] LSEG Workspace. Silver supply and demand balance (commodities research), 2021 to 2025, updated 12 May 2026 [Data export]. Cited for: silver mine production, demand, solar demand, average prices and the yearly shortfall, 2021 to 2025.

[25] LSEG Workspace. LME copper price and global manufacturing PMI, year-over-year change, to August 2026, updated 1 October 2026 [Data export]. Cited for: the change in the copper price over the year to August 2026.

[26] World Bureau of Metal Statistics, via LSEG Workspace. Copper supply and demand, January to July 2026 (September 22, 2026) [Data export]. Cited for: the refined copper surplus from January to July 2026.

Every data figure traces to a source above. Figures not represented in these sources have been removed rather than asserted.

Produced by Keaney Financial Services Corp · October 5, 2026