By Chris Puplava
Chief Investment Officer, Financial Sense® Wealth Management
September 30, 2022
There are two main approaches to investing that create a portfolio from opposite directions: bottoms-up or top-down. Bottoms-up investing is an approach that focuses on individual stocks and the health of the company while ignoring what is going on in the economy and markets. A top-down approach starts from assessing the economy from a 30,000-foot level and designs a portfolio that best suits the current and projected economic and market environment, placing less emphasis on company-specific information. Our shop has always been a top-down investment firm where we not only look at near-term trends in the economy and markets but also attempt to identify secular trends that can take place over years and decades.
The reason we use the top-down approach is that the returns on various asset classes vary widely depending on where you are in the business/economic cycle. There are four phases of the cycle: three describe the stages of an expanding economy—early, mid, and late—while the fourth phase is recession. Based on a historical analysis conducted by Fidelity, the chart below shows the returns during each of these four phases for stocks, bonds, and cash. Historically, stocks outperform all other asset classes in the early and middle phases of the business cycle (note: while not shown, commodities tend to be the best performing asset class late in the cycle when inflation is at its highest). During recessions, bonds tend to have the strongest returns, followed by cash, while stocks tend to perform the weakest.
While a business cycle is typically measured in years, secular cycles are much longer and can contain several business cycles within them. Getting the secular cycle correct also helps with long-term investment performance by identifying the top performing asset class over one or more business cycles. For example, in the 2000s, our firm moved away from a focus on technology stocks and US assets to buying foreign bonds, foreign stocks, and commodities. We did this because from a top-down, macro perspective, it was clear there was a longer-term secular shift underway from a strong US Dollar (USD) regime to a weak USD regime, and from overvalued paper assets to undervalued real/tangible assets. This secular trend was in force until it began to reverse between 2008 - 2011, which was then followed by another long and protracted secular bear market. We believe that bear market decisively ended in 2020, which marked an all-time low for oil prices at NEGATIVE $40/barrel in April of 2020 with the shutdown of the global economy.
Given the extremely depressed prices in oil and commodities that year and our forecast that much higher inflation would result from combined stimulus and suppressed supplies from lockdowns, we noted in Q3 2020 (Returning to Ground Zero) that a “Generational Buying Opportunity in Commodities” was likely setting up. Since that time, we have had above-benchmark allocations to commodities and commodity producers, and we continue to have an overweight to commodities today. While our secular view on commodities has not changed since 2020, what has is our view of the economy.
Our firm uses leading economic indicators (LEIs) to help gauge where the economy is headed so we can shift our investment focus across and within various asset classes, such as the maturity length for a bond portfolio. It is important to know that short-term interest rates are influenced heavily by the Federal Reserve’s interest rate policy. While the Fed influences long-term interest rates as well, economic trends tend to influence their direction even more. Therefore, long-term rates often fall near the end of a Fed tightening cycle even while the Fed is hiking rates as bond investors seek safety from the rising risk of recession.
We can easily illustrate how long-term yields track the economy more than Fed policy by looking at most of the last decade when rates were at or near 0% and yet long-term rates fluctuated widely. Shown below is the 10-year US Treasury (UST) yield in red alongside the ISM Manufacturing Purchasing Managers Index (PMI). The ISM PMI not only gauges manufacturing activity but also leads growth trends in the economy.
In our first quarter 2019 newsletter, we commented how the US economy was continuing to decelerate based on our reading of the LEIs as well as our own indicators. A short snippet of our January 2019 newsletter is provided below:
2019 Outlook - Buckle Up!
“Our leading economic indicators show growth in the U.S. is likely to continue slowing throughout the first half of 2019. Our firm’s interest rate momentum index leads the Conference Board’s Leading Economic Index (LEI) by one year and suggests the Conference Board’s LEI should fall well into the summer.”
Based on our view that LEIs would follow our own long-leading economic indicators, we forecasted that interest rates would fall and positioned accordingly by purchasing the iShares 20+ Year Treasury Bond ETF (TLT) for the bulk of our investment objectives that month. In October of 2019, we saw the economic landscape changing as the Fed had started to expand its balance sheet again and felt economic growth was bottoming so we sold our TLT position and transitioned portfolios for a rising interest rate environment for a double-digit return.
For the last two years, we have kept our bond allocations low and also maintained bonds with short maturities given our view for higher inflation and interest rates. While economic growth as seen by the ISM PMI jumped in 2020 and went on to reach its highest level in over a decade, interest rates were subdued and did not rise as much as the PMI would suggest as the government and Fed kept stimulating even after the brief 2020 recession was over, which suppressed interest rates (yellow region below). Economic growth rates peaked in the summer of 2021 which would argue that interest rates should start to fall but have instead unhinged from economic activity (see red bar below).
By late spring to early summer, we continued to forecast that economic growth would continue to decelerate with the odds of an eventual recession rising. The average return for long-term USTs around recessions is 28% over the last six recessions going back to 1975 (as far back as we have data), when measuring performance 12 months before and 24 months after the onset of a recession. Long-term USTs were up in five of those recessions for an average gain of 34%, while the one negative return came with the 1980 recession for a loss of -3%.
Our view this summer was that we could be staring at a recession by late summer to early 2023 which, if correct, meant that interest rates should be peaking and bond prices should be bottoming. Therefore, we began to scale into TLT with several purchases to hedge clients against the risk of a recession and decline in interest rates. In our last client letter, we remarked that the current rate of decline in the Conference Board’s Leading Economic Index (LEI) argued that it may fall below a key threshold in August:
A decline below the -1% threshold would be a significant development since it typically acts as a major demarcation line for recession. To put it more precisely, the U.S. economy has NEVER skirted a recession (going back to 1960) once the LEI’s growth rate slips below -1%. Looking at prior lead times to recession, the median over the last seven recessions (not including 2020) was 3 months (two had no lead time and one had 1 month). So, should we fall below -1% in August, we could be looking at the start date of a recession occurring between August to November 2022.
While we did not see the LEI fall below -1% in August, it fell right to it last month. The lowest value seen without a recession was in 1967 when it fell to -0.72%, so we are already below that level and likely will continue to slide as economic data continues to soften from here, increasing the odds of a recession.
When analyzing the returns on long-term USTs around recessions, we found that interest rates peak and bond prices bottom roughly a year and a half before the recession began. This is exactly what we saw ahead of the 2007-2009 recession where long-term yields peaked in the middle of 2006 and plummeted during the recession which witnessed the Bloomberg US Long Treasury Total Return Index put in an impressive 51% move during the period and rallied just under 38% around the 2001 recession.
End of Multi-Decade Bond Bull Market?
As we mentioned above, as top-down investors we not only track the business cycle but we also study long-term secular trends as they often define which assets have superior performance over others. In terms of secular trends, one of the longest on record has been the decline in long-term interest rates since 1981, which has defined the careers for investment professionals that are nearing retirement, as well as those who entered the field and are not yet even in their forties. There are very few investment professionals who were working in the 1960s and 1970s today who remember what a secular rise in interest rates looks like, but one who does and correctly called the secular turning point in inflation and interest rates is A. Gary Shilling, who wrote a book in 1983 titled, “Is Inflation Ending: Are You Ready?”
Gary is a regular guest on our Financial Sense Newshour podcast and a point he has made regarding bonds is that they have provided comparable returns to stocks with a much smoother ride. For example, from September 1981 to March 2020, the Bloomberg US Long Treasury Total Return Index provided an annualized return of 10.31% a year while the S&P 500 had an annualized return of 10.76% a year. Over this period, the S&P 500 had seven declines of 20% or more, lost a third of its value 4 times, and lost half of its value twice. Over the same time, the Bloomberg Long Treasury Total Return has been in only two bear markets with the present being the second.
Secular trends in interest rates often last 20-40 years and while secular peaks often occur over brief time periods of months to a year, secular bottoms in rates often can occur for years and even beyond a decade. You can see these trends in interest rates looking at the figure below of the 10-year US Treasury yield going back to 1900. Notice that the decline heading into WWII bottomed over the course of 13 years before beginning a nearly three decades rise into the 1981 peak. We entered another bottoming process beginning in 2012, as rates have more or less gone sideways since, then outside of the 2020 COVID crisis that briefly broke below this trading range.
The sharp rise in interest rates last month not only broke this 10-year trading range but it also broke the three-decade declining trend in interest rates that argues the secular decline in interest rates since 1981, as well as the trading range that began in 2012, has been officially broken. Prior to last month there was no guarantee that interest rates would break out to officially end the great bond bull market of the last nearly half century, particularly as U.S. recession risk was rising.
Investors Are Paid to Adapt, Not to Forecast
Louis-Vincent Gave is another prominent market strategist we have had on the show many times who frequently reminds market participants that “investors are paid to adapt, not to forecast.” Being a top-down investor, we do project where we think the market and economy is headed and invest accordingly but we also must adapt our thesis to incoming data and adjust when necessary. While our firm has been bullish on commodities since 2020 and warned that the inflationary pressures we were witnessing last year would not be transitory, we did not know how high inflation or interest rates would spike this year and that we would finally be exiting the bottoming process in interest rates since 2012. Now that we do, we need to adapt by revisiting the 1960s and 1970s, which witnessed rising interest rates, rising inflation, and rising commodity prices to study how the secular trend influenced asset classes.
During the secular decline in interest rates that began in 1981, for the four recessions that occurred since then, long-term US Treasuries bottomed on average 20 months ahead of the start of a recession. However, if we look at the three recessions that occurred between 1970 and 1983, we find that bonds did not rally well before the recession began but just as it started. Looking at the three recessions that occurred (1973-1975, 1980, and 1981 recessions) showed that the 1980 and 1981 recessions saw long-term USTs bottom during the same month the recession began while the 1973-1975 recession bonds bottomed five months prior to the onset of recession.
When we look at how long-term USTs performed six months prior to the onset of recession in the 1970-1983 period, they fell on average 10.3%, which is approximately what we are down in TLT for most of our clients. As mentioned above, based on the Conference Board’s LEI hitting -1% for August, the median lead time to recession was 3 months while two cycles had no lead time and another had one month. Thus, it’s quite possible the recession starts between August to November, and if we are correct in our recession call, that means that interest rates should be in the process of peaking, and bonds prices in the process of bottoming.
Even during a long-term secular rise in interest rates, there are periods of time in which bonds performed very well, just as during a long-term secular decline in interest rates, there are periods in which bonds did poorly. During the secular rise in interest rates seen in the 1970s to early 1980s, while bonds did not bottom far in advance of a recession, they still performed well during the recessions and served as great hedges to protect investors. For the three recessions that occurred during this period, the Bloomberg Long US Treasury Total Return Index averaged a 32% return from the onset of recession to 24 months afterwards as bonds continued to rally even after the recession had ended. For example, during the 1973-1975 recession, the Bloomberg Treasury Index rallied 40% into a peak late in 1976 as seen below.
The mildest rally occurred around the 1980 recession with a 23% gain, while the 1981-1982 recession saw the Bloomberg index rally nearly 70%.
With the full benefit of hindsight, we were clearly early in buying our long-term UST position as we were using the timing of a peak in interest rates relative to recessions that occurs during secular declines in interest rates. There is little to no lead time between a peak in interest rates and a recession during a secular rise in rates. We purchased long-term USTs to hedge against the risk of recession, so getting the recession call is crucial for us. While some have cited the strong jobs market as evidence that the US is not in danger of falling into a recession anytime soon, a history lesson would argue against that.
Just as bonds performed differently during a secular rise in interest rates than they do in a secular decline, so do economic trends. Part of the reason the Fed continues an on aggressive hiking campaign, is that they see a robust jobs market that they use to justify pressing onwards to squash inflation. Looking at the 1990, 2001, and 2007-2009 recessions, US payrolls peaked on average one month prior to the onset of a recession and so by the time payrolls peaked, the US was virtually in recession and, thus, looking at payrolls would lead the Fed astray.
This is even more so during periods of high inflation and a secular rise in interest rates. Looking at the four recessions between 1970 to 1982, showed that payrolls peaked on average between 3-4 months into recession, so it would not be uncharacteristic for the recession to have started even while the economy continues to add jobs.
Considering our belief that our recession call remains valid, we have held onto TLT during this difficult decline as we did not want to sell into panic, let alone after the worst decline ever. Since the peak last year, the Bloomberg Long-term US Index is down 34%, the largest decline in nearly a half century. Selling now would be akin to selling stocks in March of 2009, which was the worst decline in a half century for stocks.
We have spent the bulk of this client letter to directly address the elephant in the room, which has been our move into long-term USTs (TLT) to hedge recession risk and why we still hold this position. We will cover the rest of the portfolio and our outlook in the next client letter. Should you have any questions, please do not hesitate to reach out to your wealth advisor.
Advisory services offered through Financial Sense® Advisors, Inc., a registered investment adviser. Securities offered through Financial Sense® Securities, Inc., Member FINRA/SIPC. DBA Financial Sense® Wealth Management.
Copyright © 2022 Chris Puplava