When The Six Million Dollar Man first arrived on television in 1973, six million dollars bought you a bionic former astronaut who could outrun a car. Today, rebuilding Steve Austin would cost about $45 million. Put the other way around, six million of today’s dollars buys roughly $800,000 of 1973 purchasing power.
And even then, you would probably only get a feasibility study, an environmental review, and a consultant’s invoice. Well, that’s progress, of a sort.
Nobody devalued Steve Austin; they devalued the dollar. Everything that follows is a footnote to that distinction.
“A Man Barely Alive”
Prices move all the time, and most of those movements are useful. When wheat rises after a poor harvest, that tells consumers to economize and farmers what to plant next season. Relative prices are the nervous system of a market economy. They are supposed to twitch.
General inflation is different. It is the unit of account losing purchasing power. When inflation is higher than expected, it shifts value between people, usually from those holding cash and fixed nominal promises toward those owing fixed nominal debts. Benjamin Franklin understood this in 1780, having helped cause it. Writing about the collapse of Continental currency, he described depreciation as a tax on money holders, proportionate to how much they held and for how long.
The United States government is the largest debtor in the developed world, with gross federal debt north of $39 trillion and annual interest expense above $1 trillion.
Inflation can reduce the real burden of existing fixed-rate debt, but it is not a magic eraser. The arithmetic also depends on the average rate paid across the debt stock, how quickly it must be refinanced, nominal GDP growth, and, above all, whether the government continues to run large primary deficits. Devaluation helps the debtor. It does not repeal accounting, and it is unlikely to improve sentiment among the population forced to endure it.
“We Have the Technology”
The Consumer Price Index is competently and honestly constructed, but it also answers a question many households did not realize had been asked.
The BLS is not simply pricing an eternally fixed shopping basket. It is trying to approximate the cost of maintaining a broadly comparable standard of living. Those ideas sound similar, but they are not quite the same. CPI already allows some substitution among closely related products, while chained CPI goes further by recognizing that consumers change what they buy when relative prices move. The familiar claim that statisticians simply replace expensive beef with chicken and declare the consumer no worse off is wrong.
But the broader distinction matters. A cost-of-living index is interested partly in what consumers can substitute toward. A household is often more interested in what it can no longer afford.
Housing makes the point rather neatly. Before 1983, US inflation measures treated owner-occupied housing more like an asset purchase, incorporating house prices, mortgage interest, and related costs. Owners’ equivalent rent replaced that approach: an estimate of the housing service homeowners receive from their property.
Economically, that is defensible. A house is both somewhere to live and an asset, and asset prices do not normally belong directly inside a consumer price index. Psychologically, it can be baffling. A family trying to buy its first home does not experience the housing market as an imputed rental service. It experiences a deposit, a purchase price, and a mortgage payment.
In 2024, Larry Summers and his co-authors reconstructed an alternative inflation measure incorporating mortgage rates and other borrowing costs. Official CPI peaked at 9.1%; their broader measure approached 18%. They did not prove that official inflation was fraudulent. They showed that borrowing costs explained much of the gulf between relatively strong conventional economic data and extremely weak consumer sentiment. Americans were not necessarily confused. They were reacting to costs that the headline inflation rate was never designed to measure.
The same applies across households. A retired homeowner, a young renter, a commuter, and a family refinancing a mortgage can experience very different inflation while the BLS measures the aggregate perfectly honestly.
There is a thriving industry selling supposedly “true” inflation numbers, and most of it should be treated with caution. The BLS isn’t making the numbers up; it’s just that no single inflation measure can describe everybody’s economic life.
Indeed, the famous Boskin Commission reached almost the opposite conclusion in 1996: that CPI tended to overstate increases in the cost of living because it was too slow to recognize substitution, falling prices, and improvements in quality. This isn’t a conspiracy. For an investor, it is basis risk.
Social Security payments are indexed to CPI-W. The Federal Reserve targets PCE, which has broader coverage, different weights, and a much smaller housing component. US tax brackets use chained CPI. TIPS are indexed to CPI-U. Each measure has a technical rationale.
For an investor, the important question is which ruler is being used, and whether it resembles the liability you are actually trying to protect.
“Better. Stronger. Faster.”
Steve Austin’s decade is the one everybody invokes when discussing inflation, but almost nobody examines it carefully.
From the market peak in 1966 to the early 1980s, the S&P 500 price index went broadly nowhere and lost around two-thirds of its purchasing power. Reinvested dividends rescued the result from disaster, turning an enormous real loss into something closer to prolonged real stagnation.
But that rescue came from a very different starting point.
The S&P 500 yielded roughly 3.6% in 1966, and dividend yields later rose above 5%. Today the index yields barely more than 1%. Sixteen years of violent volatility for little or no real return was unattractive even with a substantial stream of cash income. Repeating the exercise today would require much more help from earnings growth, valuations, and capital allocation.
Companies now return much more cash through buybacks. In the 12 months to September 2025, S&P 500 companies spent roughly $1 trillion repurchasing shares, against about $665 billion in dividends. That money matters, but a buyback is not a dividend wearing more fashionable clothes.
Repurchases are discretionary, concentrated among the largest companies and partly offset by shares issued to employees. They create value only where the share count genuinely falls, and management has not overpaid. They also provide no cash to the shareholder unless he sells part of his holding.
Tax matters too. In 1966, dividends were generally taxed as ordinary income and the top statutory federal rate was 70%. Today qualified dividends receive preferential rates, while buybacks allow tax to be deferred until shares are sold. Historical comparisons therefore need to be made after tax as well as after inflation.
None of this means equities are doomed to repeat the 1970s. It does mean that relying on a broad equity index as an automatic inflation hedge looks optimistic from today’s starting point. That is my best British understatement.
Long Treasuries offered no such ambiguity. Yields rose from around 4% toward the mid-teens, crushing existing bonds as surely as a punch from Steve’s bionic arm.
Treasury bills fared better before tax, but even they could produce deeply negative after-tax real returns. Earn 11% while inflation is 11% and you have made nothing in real terms. The tax authorities, however, regard the entire coupon as income. You were taxed on your own erosion, and they say the government doesn’t have a sense of humor.
Gold was the outstanding asset of the period, rising several hundred percent in real terms before surrendering much of that advantage during the following two decades. If the CIA had put the money into gold instead of Steve, it would be worth more than $300 million now.
Gold benefited from inflation and monetary disorder, but also from the release of a price that had been politically constrained for decades. It has useful monetary qualities. Consistency is not one of them.
There is one final historical complication. The inflation index of the 1970s was constructed differently. Mortgage interest sat directly inside it, so higher interest rates intended to fight inflation could themselves push measured inflation upward. On modern methodology, the peak would have been materially lower.
Methodology, taxation, starting valuation, and the way companies distribute cash can all change the investment result. Which brings us to the practical question: What liability are we trying to protect, and which asset actually resembles it?
“Gentlemen, We Can Rebuild Him”
TIPS did not exist in 1973. That was unfortunate for investors then and potentially useful for investors now. Recent 10-year TIPS have offered real yields of roughly 2.4%, while 30-year real yields have approached 3% (as of August 7, 2026).
Ten-year breakeven inflation has been around 2.3% (as of August 7, 2026): broadly, the average CPI inflation rate at which a nominal Treasury and comparable TIPS should produce the same return if both are held to maturity. There are two important wrinkles. First, the Fed targets PCE while TIPS compensate investors for CPI. The measures are not identical, and CPI has generally run somewhat higher. So, a 2.3% CPI breakeven does not mean the market expects the Fed to miss its 2% inflation target by 30 basis points.
The investment decision is simpler. The nominal Treasury holder accepts the risk that CPI averages more than roughly the breakeven rate. The TIPS investor receives more than 2% in real yield and does not have to take that particular bet.
The Fed’s record is not one of continuous 2% inflation. Inflation spent years below target before the pandemic and substantially above it afterward. Averaging the two periods makes the result look considerably tidier. They do not cancel for the saver.
The Fed targets the future inflation rate, not the price level. Once purchasing power has been lost, returning inflation to 2% does not restore it. It merely promises that what remains will erode more slowly.
At today’s real yields, the question is therefore less whether an investor can forecast inflation to the second decimal place than how much he is being paid to retain the risk. And that leads to the portfolio conclusions.
The second is less technical but far more important. Inflation protection is not duration protection. In 2022, inflation reached 9%, yet many TIPS funds still lost money because real yields rose sharply. A long-dated inflation-linked bond remains a long-dated bond: its principal may adjust with CPI, but its market value can still fall heavily when real rates rise for investors with known liabilities, that argues for paying as much attention to maturity as to inflation protection itself.
What Should an Investor Actually Do?
- Start with the liability, not the inflation forecast. TIPS hedge official CPI. They do not perfectly hedge a client’s healthcare, rent, insurance, education, or care costs. The relevant inflation rate is the one attached to the spending you are trying to fund.
- Real yield matters. TIPS yielding CPI plus more than 2% are a very different investment from TIPS yielding CPI minus 1%. The security has not changed. The price has.
- Do not confuse inflation protection with duration protection. Long-dated TIPS can lose substantial money when real yields rise. That is what happened in 2022. An inflation-linked bond remains a bond.
- Match maturities where possible. Individual bonds or ladders aligned with expected spending reduce the danger of accidentally turning an inflation hedge into a large duration position.
- Treat breakevens as prices, not prophecies. The useful question is not whether 2.3% is the “correct” inflation forecast. It is whether the compensation for taking the other side of that inflation risk is attractive.
- Equities are not automatic inflation hedges. Starting valuation, dividend yield, taxation, and capital allocation matter. The experience of the 1970s is considerably less comforting when today’s starting yield is barely 1%.
- Think after tax. A nominal return that keeps pace with inflation before tax can still destroy purchasing power after tax.
- Use real assets as complements, not magical inflation coupons. Energy, infrastructure, farmland, utilities, and gold can help match particular liabilities, but they introduce equity risk, valuation risk, regulatory risk, liquidity risk, and basis risk of their own.
- Above all, know which ruler you are using. CPI, PCE, and personal inflation can tell different stories. A perfectly constructed index is still the wrong hedge if it measures somebody else’s liabilities.
They told us they could rebuild him, and that they had the technology. They rebuilt Steve Austin once, but after the gold link was broken, they never bothered rebuilding the dollar. With the 1973 dollar retaining only around 13 cents of its purchasing power, investors need to avoid the same fate for their capital.
References
Alphaville. 2024. “TIPSplaining a Lousy Inflation Hedge.” Financial Times. January 30.
Bolhuis, M., J.N.L. Cramer, K.O. Schulz, and L.H. Summers. 2024. “The Cost of Money is Part of the Cost of Living: New Evidence on the Consumer Sentiment Anomaly.” NBER Working Paper 32163, February.
Boskin Commission Report. 1996. “Toward a More Accurate Measure of the Cost of Living.” December 4.
Bureau of Economic Analysis. 2010. “What Accounts for the Differences in the PCE Price Index and the Consumer Price Index?” November 3.
Cochrane, J. 2024. “Why Inflation Still Looms Large for US Voters.” Chicago Booth Review, May 10.
Debrett’s. “The Art of Understatement.”
Franklin, Benjamin. 1780. “Benjamin Franklin to Thomas Ruston.” Founders Online. October 9.
Horstmeyer, D. 2025. “Inflation-Protected Bonds Fail a Key Test: They Don’t Help When Inflation Is High.” Wall Street Journal, November 13.
Palmberg, J. 2026. “Gold Market Commentary: Making Waves.” World Gold Council, August 6.
U.S. Bureau of Labor Statistics. 2019. “Common Misconceptions about the Consumer Price Index: Questions and Answers.” August 15.
Wikipedia. “The Six Million Dollar Man.” Last modified August 9, 2026.