Inverted Yield Curve

U.S. Treasury Bond, 1977, US$ 1 millionThis is a fairly geeky financial technical analysis post.  I usually post such material as Gnome-o-Grams on my own Web log, but as an experiment, I’m posting this one here to see if there’s any interest among the membership or wish to see further posts of this kind.  If you’d like to see more like this here, please indicate by liking this post or commenting, including topics you’d like to see discussed (after reviewing those already published, linked above).  As always when anything discussing finance or investing, I don’t make recommendations; you’re entirely responsible for your own decisions and would be crazy to interpret anything I say as a course of action you should pursue without coming to your own independent decision.

The yield curve is a measure of the sentiment of investors, particularly conservative investors who own fixed-income securities (bonds and equivalents).  (Much of the financial press covers the equity [stock] markets and neglects the bond markets, but the bond markets dwarf the stock market in valuation.  Bonds aren’t [usually] exciting [and when they are things are generally unpleasant], so they don’t get much attention, but if you’re interested in the flow of funds [and you should be], that’s where you ought to be looking.)  The yield curve simply plots, for equivalent fixed-income (debt) securities (bonds), the relationship between the yield of the security and the time to its maturity (when the investor gets his or her money back).  For example, consider the most widely traded securities in the world: U.S. Treasury debt.  These instruments have various names: Treasury Bills, Treasury Notes, and Treasury Bonds, depending upon their time to maturity, but they all are obligations of the U.S. Treasury and bear the full faith and credit of the United States.  They are considered as close to risk-free as any paper investment in the world.

In normal financial circumstances, which is almost all the time, the longer the time to maturity of the investment, the greater the yield.  For example, in May 2018, a 90 day U.S. Treasury Bill yielded around 1.6% interest (per annum, as will be  all figures quoted).  If you were willing to lock in your money for a year, you’d get about 2.25%.  Go out five years in a Treasury Note and that would go up to around 2.8%, and with a ten year commitment you’d get 3%.  Longer terms would get higher yields, but it’s asymptotic—at thirty years in a Treasury Bond you’d only get 3.2%.

Although nothing in investing can be considered normal since 2008, and yield curves in the era of sound money and before artificial repression of interest rates were more linear, the slope of this curve is representative of normal conditions.  An investor who locks up their money for a longer term at a fixed rate of return is assuming the risk that they may be repaid in inflated money worth less than that they used to purchase the bond, or that they may forgo more attractive investment opportunities during the time their money is locked up in the bond.  (You can always sell a bond before its maturity, but if interest rates have risen since the time you bought it, you’ll take a loss compared to what you paid as the purchaser will discount the price so the bond yields a market rate of return.)  Therefore, they usually demand a “risk premium” in the form of a higher interest rate on the investment to compensate for these risks.  There is always some rate at which investors judge worth tying up their money for a longer time.  The greater the perception of risk, the steeper the yield curve.  Thus, the yield curve is a sensitive indicator of investors’ perception of risks in the future.

Now let’s look at the yield curve, as measured by the spread (difference in yield) between the 10 year U.S. Treasury note and 90 day U.S. Treasury bills, plotted between 1982 and the present, courtesy of the Federal Reserve Bank of St. Louis FRED database.

FRED: Yield curve, 1982 to present

Note how the yield curve, measured by just these two points on the maturity of a single debt security, has varied over the years.  At highs, it approaches 4%, while at lows—now that’s interesting—it goes negative.  What could that signify?

Well, at the simplest level, it means that investors are willing to accept a lower yield for locking their money up for ten years than simply parking it with the Treasury for 90 days.  Are they crazy?  What are they thinking?

This phenomenon, when the yield curve goes negative, is called an inverted yield curve.  There are a number of possible explanations for it, but most of them do not bode well for the economy in the near future.  If an investor anticipates an economic slowdown, that slowdown is likely to reduce the demand for capital and will be reflected in lower interest rates.  Buying a longer-term bond allows “locking in” today’s higher interest rate for the full time until the bond matures, regardless of where interest rates go during that period.  If the investor stayed in short-term instruments, they would be rolled over periodically during that time and, if interest rates had fallen, would be renewed at lower yield, resulting in less income compared to the long-term bond.  This preference causes investors to bid up the price of longer bonds compared to those with shorter terms, resulting in an inversion of the yield curve.  In addition, investors anticipating the sharp drop in equity (stock) markets which often accompanies an economic downturn, may prefer to park their money in a risk-free bond which will provide a guaranteed stable return throughout the anticipated time of turbulence.  Thus the yield curve reflects investor sentiment: when it is steeply positive (long rates well above short rates), investors are generally optimistic about the future and demand compensation in the form of higher interest rates for locking their money up and forgoing other opportunities.  When they’re less sanguine about what’s coming, the prospect of not only a guaranteed rate of return on their money but also return of 100% of their money when the bond matures sounds like a pretty good deal compared to the alternatives (such as buying into a historically overpriced stock market late in an economic expansion cycle) and they’re willing to accept a lower rate of interest to secure that return.  (If you think this paragraph was tangled, check out the work of academic economists on the expectations hypothesis, which says more or less the same thing in intimidating mathematics.  There are a number of competing theories to explain the yield curve, and, as usual, economists differ on which best explains the phenomenon.)

Whatever the motivations of investors which result in the rare occurrence of an inverted yield curve, that circumstance has been a reliable indicator of bad times just around the corner.  Since 1970, shortly after the dawn of the era of pure paper money and the inflation and exchange rate instability it engendered, there have been eight periods where the yield curve went inverted based upon the monthly rates of 90 day and 10 year U.S. Treasury debt.  Seven of these eight periods of yield curve inversion have been followed by economic recessions as declared by the National Bureau of Economic Research (NBER).  The time between the onset of the yield curve inversion and the start of the recession has varied between 6 and 17 months, but a recession has always ensued.  (Recessions are marked by the grey bars in the yield curve chart above.)  Further, there has not been a single recession during that almost half century which was not preceded by an inversion of the yield curve.

“But, you said, ‘Seven out of eight times.’  What about the eighth?”

That’s where it gets interesting, and newsworthy.  After ten years in positive territory, the yield curve went negative in May, 2019 and has remained negative since then.  If a recession does not follow this signal, it will be the first time it has failed to forecast a recession since 1970.  An inversion of the yield curve does not forecast the date of onset of the recession or its severity, but history advises that if you’re willing to bet a recession (with all of its sequelæ for the stock market, unemployment, budget deficits, and politics) will not start within 18 months or so after the inversion of the yield curve you must really believe that “this time is different”.  That, of course, is what all of the sell-side analysts are telling you, but listen to them at your own risk.

My guess, and it’s only a guess, is that there may be some financial turbulence ahead, including a media-hyped (but long-overdue) recession in the run-up to the 2020 elections in the U.S.  This will, of course, be presented as evidence of the “failure of capitalism” and reason for the electorate to vote for whatever slaver nostrums are on the menu in that contest.  There are other indications, still ambivalent, of a gathering storm: gold has just in the last week blown through its five year resistance level at US$1350/troy ounce and raced to more than US$1400, and Bitcoin has exploded to around US$11000/BTC.  These are, no doubt, driven in part by fears of the war the Deep State is trying to gin up between the U.S. and Iran in order to take down Trump, but they may also limn the first signs of the inevitable consequences of the market’s realisation that there is no consensus in Washington to do anything about the runaway deficits, debt, and inevitable insolvency of the U.S. government and the reserve currency it prints.

What is clear is that inversion of the yield curve is one of the most reliable indicators over the last fifty years that a period of “business as usual” is coming to an end.  The indicator has now given its signal.  It is prudent to consider the consequences and ponder how prepared you are for what may come next.


Author: John Walker

Founder of, Autodesk, Inc., and Marinchip Systems. Author of The Hacker's Diet. Creator of

11 thoughts on “Inverted Yield Curve”

  1. John Walker:
    If you’d like to see more like this here, please indicate by liking this post or commenting

    Yes, please!


    I’d especially like to see an analysis of the long-term inter-generational wealth transfer patterns surrounding Social Security.  I once went surfing on the Bureau of Labor Statistics website to see if the necessary grouped aggregate data was available — counts and sums of FICA receipts vs. eligible income broken down by age and income percentile per year — with no such luck.  I suspect it would take a formal request to BLS with research $$ to get that information.


    If the inverted yield curve were a reliable sell signal, I have to wonder why it’s not already priced into the market. After all, the signal has been there for over a month, which is plenty of time for traders to decide that it’s real and act accordingly. This is not to say that the signal is wrong this time.

    Last recession, it took about 1.3 years between the dip in the yield differential and the peak of the S&P (July 06 to Oct 07). In the previous recession it was only one month (July 2000 to Aug 2000). It’s hard to find anything very interesting happening to the US equity market for the 1989 dip in the yield curve. The sometimes lengthy and widely varying gap between effects in these two markets (debt and equity) makes me wonder about the underlying causes of the behavior of these two indicators. 

    S&P for the same period as the yield differential curve in the OP is plotted below on a log scale. There year 2000 and year 2008 market dips follow yield curve inversions but the 1989 inversion has no corresponding equities response. On the other hand, the 1987 minicrash is not preceded by an inversion in the yield curve. By my reckoning, the predictive value of the yield curve is two out of the last four, at least when it comes to US equities. 

  3. The book I am reading, “Risk Savvy”, proposes a plan for investment that is simple. You find what you feel are good investments and divide your money equally among them. This strategy has succeeded over more complicated ways. One of the points of the book is complicated methods sound good but if the system is uncertain they don’t work so well.

    I myself don’t much about investing. I know the old joke, “If you want to make a small fortune start with a large fortune.” though.

    Anytime I get to read John Walker’s insights is good. I learn.

  4. There is massive turbulence ahead, but it will be in the switch from a global finance driven economy back to a  hard asset nationalistic one.

    Investments in the wrong countries and currencies will cause uncertainty.

    The   post WW2 era is finally coming to a close. If you bet on continued globalization, repressed wages and easy appropriation of intellectual property, cartel like pricing in agriculture and such, it will be ugly.

    The ride will be rocky until asteroid exploitation kicks in and changes all commodity prices without mercy.

  5. drlorentz:
    By my reckoning, the predictive value of the yield curve is two out of the last four, at least when it comes to US equities.

    Inversion of the yield curve is not a particularly useful indicator of future performance of the stock market, except perhaps indirectly.  It has been, however, for the last fifty years, a very good leading indicator of the onset of recessions.  Ignoring the inversion which just started last month, since 1970 every inversion has been followed by an NBER-declared recession within the next 18 months, and no recession has occurred without a preceding inversion within that interval.  Now, a sceptic might say, “Well, that isn’t useful if you don’t know how long it will be from the time of the inversion until the onset of the recession.  If the two were completely uncorrelated it’s still likely an inversion will happen sometime before a recession.”  But what makes the signal useful is that recessions so far have always occurred much sooner after the inversion than random timing would produce (see the chart in the main post).  Further, there’s a consistent pattern: a fairly steep V-shape around the inversion, with a recovery underway just about when the recession is declared (declaration of recessions tends to be a trailing indicator).

    As you noted, the stock market is only indirectly correlated to the business cycle.  While you’re more likely to make money in the stock market during an economic expansion than a contraction, the timing is imprecise and difficult to forecast due to the market’s discounting of anticipated future events.  The 1987 crash (I was chairman of an S&P 500 company during that event, and it didn’t seem like a mini-crash to me) was an example of an unexpected stock market wipe-out in the middle of a robust economic expansion which was, in retrospect, nearer the mid-point between two recessions than either one around it.  Equity markets are prone to irrational exuberance, panics on both the up- and down-side, and unwarranted despair.  Debt markets are more directly coupled to the flow of funds in the economy and produce signals worth paying attention to, even if they’re not particularly useful for stock market timing (which may, in fact, be impossible if you subscribe to the efficient market hypothesis).

  6. 10 Cents:
    The book I am reading, “Risk Savvy”, proposes a plan for investment that is simple. You find what you feel are good investments and divide your money equally among them.

    This is more or less what wealthy people have been doing for centuries.  You calculate an asset allocation among stocks, bonds, and cash based upon your age, assets, amount of savings you’re planning to invest each year, then you maintain that allocation, adjusting it every decade or so as you age and pass through different phases of life.  (For example, when you’re younger and have a high income, you can tolerate more risk (greater stock allocation) with a goal of better long-term return—if the market tanks, you have plenty of time to wait for it recover, which an older person might not.  Older people and retirees will prefer an allocation more heavily weighted toward bonds because preserving their capital (for their own use and to pass on to heirs) and generating income to live on outweighs the potential benefits of assuming more risk.

    Here is a free asset allocation calculator which generates  portfolios for various investor profiles.  While there are a lot of different approaches and subtleties (for example, the mix of stock sectors and bond maturities and diversification across industries, countries, and currencies), most financial planners will come up with something similar.

    With an allocation like this you don’t try to time the market or adjust your allocation based on short-term conditions or forecasts.  Neither do you try to pick stocks or trade in and out of the market.  Just buy a broad-based index fund (for example one that tracks the S&P 500 or equivalent for markets in other countries) and re-invest the dividends in the fund.  Most professional stock pickers and market timers underperform the broad averages over periods of 10 years or more.  Do you think you can do better than they do at their own game?

    Every now and then (say, yearly or biennially), re-adjust your portfolio to the chosen allocation by selling over-weighted assets and moving the proceeds to under-weighted ones.  For example, suppose your allocation is 60% stocks, 30% bonds, and 10% cash, and due to a booming stock market, you’re now at 70% stocks with bonds and cash now accounting for 30% rather than the 40% you planned.  To re-balance the portfolio, sell part of your stock fund and move the money into the bond and cash sectors to restore the planned allocation.  Conversely, if there has been a stock market decline that reduces the stock fraction to, say, 50%, move money into stocks.  Note that over the long term this means you automatically “buy low and sell high”.

    Ultimately, rely upon compounding.  Every year or so, re-read Richard Russell’s Rich Man, Poor Man” [PDF], and recommend it to anybody who asks about your investment strategy, especially young people who stand to benefit from it the most.

  7. John Walker:

    The main takeaway from that graph is that the number of months since the last recession has almost no predictive value given that they span an order of magnitude. Any time-series of random numbers will continually produce new lowest and highest values. In other words, records are made to be broken.


  8. “Longest economic expansion on record.”

    I think we can attribute this to Team Obama.

    By keeping the regulatory boot on the neck of American business, they prevented the recovery that they clearly expected would happen.   This reveals that they do not understand how the economy works.

    By keeping a lid on growth by strangling business, they managed to force the recovery to happen very very slowly.

    It is only in the Age of Trump that American business has had the choker leash removed.   I expect this expansion to continue for a few more months, until a vigorous anti-economy all-bad-news-all-the-time 2020 campaign effort by mass media kills consumer confidence and triggers the big slump.

  9. John Walker:
    What is clear is that inversion of the yield curve is one of the most reliable indicators over the last fifty years that a period of “business as usual” is coming to an end.

    This was a great post and I have completely agreed with your analysis for several years.

    I’m fortunate to have invested in Google, Amazon, FB over 10 years ago but took my money out of FB and ran when it traded over $210. It had a nice day today at $194 but I’m leery as the govt is now going after MZ (for monopolization of all things!) as it did Bill Gates and the late Sam Walton. Growth stocks are relatively easy, but my husband  focuses on under the radar stocks that produce high dividends.

    We both realize however, that the market is vulnerable to so much more than the math- politics, trade wars, the horribly ignorant MSM so we have decided to convert more of our investments to cash and of course, insure it.

    It’s definitely not as much fun now. Quarterly earnings don’t seem to matter as much; other nonsensical elements do.


Leave a Reply