Thanks everybody for joining us. This is our third Gundlach Unlocked webcast where I go through some macro themes and a little bit of micro here and there, too. Interestingly, for the last three months, the S&P outperformed the NASDAQ. Actually, a 60/40 with the NASDAQ returned only about 8% if you use the NASDAQ instead of the S&P 500. So let's get started.
We have the Bloomberg Aggregate Bond Index yield to worst on the screen going back to the late '90s, and we can say that we have been very range-bound on the yield of the Agg since about four years ago—or yes, four years ago. And so the range has basically been the low 4s on the low end and something in the 5% zone, except for that one spike in 2024. We've got these dashed horizontal lines on here, and that represents the average yield. The blue line is for the last 30 years—the Agg averaged 4.03%; for the last 20 years, 3.25%. And interestingly, the last 10 years was slightly higher than the 10 years prior. So yields aren't what you would call suppressed anymore, but there is a real interest rate in the Bloomberg Aggregate, which is a good thing of course. We have a spread above that in many of our funds, so many funds are now in a 6% yield. If you take high-risk fixed income like local currency emerging market or the bank loan index, you have yields of about 7%, which seems like a pretty good competition, I would say, for a stock market that has, as we'll see, a Shiller CAPE ratio that's basically at the highest level of all time.
This is a rising long-term interest rate environment, and it has been now for six years going on seven years. We see that all of them except Switzerland have been moving up in sync. The one that's most notable is Japan, which for years was suppressed nearly zero and now it's at 3.97%, which is not very far away from now the US 30-year at 5.24%—it's not even 150 basis points. We see that all these developed countries have been moving up in lockstep, with the UK having experienced the highest increase in yield.
We see that the 30-year Treasury bond bottomed in 2020, and that also was the bottom of this channel. The red lines are two standard deviations away from that centerline. We see that we had an enormous rate rise from a low of incredibly 27 basis points at the low point in 2020 now at 5.24%, for a loss of over 50% still on 30-year Treasury bonds. We're still pretty near the high here at about 5 and a 1/4%. One of these things when you have a sideways market that goes on and doesn't really rally—see, we had this big sell-off increase in the bond yield and it didn't retrace hardly at all. Usually that means that if it can't rally to correct a huge, almost 500-basis-point rate move, it probably means that the next move is going to be a continuation of the upward trend.
I built this model a long time ago to do a starting point for where the 10-year US Treasury should be as a base case. What we have here is the 10-year Treasury yield is the tan-colored line, the model is the darker line, and the forward aspect of the model is the yellow line. What we do is we take the German 10-year bond yield and we take the 7-year average of US nominal GDP. Amazingly, it gives us a serviceable starting point for where the 10-year Treasury yield should be. You'll notice in the rectangle at the bottom, the R-squared of these two lines is an amazingly high 0.93. If you just did it from say 1990 instead of went back to 1986, it would clearly be even a higher R-squared. But right now the model says that you could expect the 10-year Treasury yield to be at 4.71%, and lo and behold, it looks like it's 4.78%. So it's very much in the context of where it should be. But again, it seems like the path of least resistance could very well be higher, and we'll see why in a few moments.
I've often talked about the 2-year Treasury yield and the Fed, and they're kind of out of sync again. We had really out of sync in 2022 where the 2-year Treasury yield was laughing at the Fed staying near the zero bound with the 2-year Treasury 200 basis points higher. That's the biggest gap in my career, which goes back over 42 years. Then we see that the Fed was overly offsides the other way in 2025. Now the Fed funds rate looks like it should probably be up about 50 basis points higher than it is right now based upon where the 2-year Treasury yield is. Fed day is next week on Wednesday and we'll see what happens. The WIRP function that measures the probability of the Fed changing rates by the shape of the yield curve says there's about a 60% chance of a Fed hike based upon the pricing at the short end of the Treasury curve. There's something about Kevin Warsh that I don't quite fully trust, and I'm leaning against the 60%, although I'm not highly convicted on this. I wouldn't be surprised if the Fed did not raise interest rates next week. If that's the case, I would expect long-term interest rates to rise fairly significantly in the aftermath of the Fed. If they do hike interest rates, well, then perhaps the bond market will stay about where it is right now.
This is something that I used last webcast. It's from J.P. Morgan Asset Management. The y-axis is the ISM manufacturing prices paid. So, of course, when prices paid go up, you would expect the Fed to be more inclined to hike interest rates rather than cut them. Then the x-axis is the ISM manufacturing employment index. And so when that is above 50, one would expect the Fed to be more likely to hike than to ease; if it's below 50, more likely to ease than hike. Then there's all these little dots scattered on the screen. The blue dots were Fed easing, and the orange-ish red dots are the Fed tightening. I drew these rectangles, adding some further information to the J.P. Morgan Asset Management exhibit. At the lower left, we see all the dots in that rectangle are blue except for three or four, and I've got arrows pointing to those three or four. Those were under Paul Volcker in early 1982, where he was absolutely not following the bond market. He was proactively and sometimes impulsively, not even on meeting dates, just out of the blue changing interest rates. One famous Saturday evening, he raised rates by hundreds of basis points; it was called the Saturday Night Massacre. Then the upper right, you see another rectangle that's drawn, and that's when you would expect to see more tightening rather than easing because you would have high prices paid, which means inflation, and you'd have high employment. So both sides of the Fed's mandate would suggest a tightening. They're all tightenings except for about four dots, and that was under Arthur Burns, who was browbeaten by the presidents at that time that he was serving for to keep interest rates artificially low. Of course, that very significantly helped usher in an era of very, very high inflation in the United States. We'll see a chart on that in a few moments.
The larger orange-ish dot that's right there to the right of the 50 vertical line and right above the 70 line on the y-axis and above the 50 line on the x-axis kind of suggests that, if anything, the Fed should be tightening and not easing interest rates. Although if you look at all the dots around that larger orange dot, some of them are red and some of them are blue, so there's no definitive takeaway for this exact stage, although I think there are a little more tightening dots than there are blue easing dots.
Now we start to see some spread changes in the bond market. Here we see that on the left-hand side of the display, the lighter blue line is the spread on investment-grade bonds in the US corporate bond market excluding the AI sector, and then we see the darker line, which is the AI sector only. One thing that's fairly clear is that the broader investment-grade bond market has not really seen any meaningful spread widening, but the AI market has seen monumental spread widening vis-à-vis the investment-grade area. We see that the AI spread was only 50 and now it's up at about 125, so 75 basis points of widening over that time period. There's no change in investment-grade spreads. On the right side of the exhibit, we do the same thing for high yield, and here it's even more dramatic because the widening of the AI sector has gone from about 180 basis points to 325 or so. So we're significantly widening, while the non-AI sector in high yield, the lighter blue line, is actually right near its tights of the year. So we've seen a big divergence here, and this is something that we should really be paying attention to. We all know that the US Treasury is borrowing way too much money with a budget deficit of up 6 or 7% of GDP, but now we have all of this demand from AI and AI-adjacent businesses, which will undoubtedly put further spread pressure on the AI sector. So I'm not really sure who's buying all these AI-related bonds. It might be insurance companies that are owned by private credit firms that are owned by private equity firms that are dictating the activities of these captive insurance companies. But it's clear that the market is having a hard time digesting this amount of supply, and the supply is going to continue to be an avalanche in the AI sector. So we have the Treasury borrowing way too much money, and we have corporations with a seemingly insatiable demand for issuing debt, and the market is clearly, using the dark lines here, starting to demand higher compensation.
I hear people talk about TIPS a lot versus nominals, and we like TIPS. We have TIPS in some of our low-risk funds. We like the short-term TIPS because we think that the implied inflation by the comparison of nominals to TIPS is implying a way too low inflation rate. They're basically implying that the Fed is going to reach their target immediately of 2% and stay there. I think that's very unlikely, and so I think TIPS are too cheap on the short end. What I'm showing on the exhibit on the screen is long-term TIPS—30-year TIPS versus the 30-year nominals. A lot of people say they like TIPS, and I've even seen some fairly frequent guests on financial media talking about how they like long-term TIPS right now because they don't like long rates with the amount of borrowing that's happening by the Treasury. But it's obvious that these lines are very, very similar. Just look at the bottom, which is the difference between the two, and we see that it's been absolutely stable for the past five years. So TIPS do not hedge you. If you don't like nominal Treasuries, there's no reason to believe that 30-year TIPS are going to protect you because they've had exactly the same rate rise since the end of 2021 as the nominal bonds. So do not buy long-term TIPS thinking that it's going to somehow hedge you if you don't like long-term 30-year nominal Treasuries.
So let's take a look at inflation. Kevin Warsh definitively stated at his last press conference that the 2.00 is their goal and that they will achieve it—he promises—and he says that they use the PCE deflator. We have on the screen here he quoted the 12-month deflator, which he correctly pointed out was 3.7, and he also went so far as to say the six-month annualized rate of change of the PCE deflator is actually higher. So it's actually been rising more in the last six months than it did in the six months prior. The PCE deflator is not getting really much better. The six-month annualized is higher than the 12-month annualized, and neither is anywhere close to 2%. Now let's look at year-over-year looking at core and headline. Here we have the core rate is 3.3 and the headline is 3.7. Both appear to be basically in a rising trend since the middle of 2024, although it has stopped rising in recent reports. It'll be interesting to see what the next inflation readings bring because that's going to greatly inform, I think, the Fed's tendencies towards future policy.
This is here just for entertainment's sake more than anything else. We have overlaid the experience of the '60s into the early '80s of the inflation rate using the headline CPI. In the '70s and '80s, we see the CPI got up to 12.5% and then eclipsed it earlier in the '80s way up at nearly 15%. And then we have the most recent rate rising experience from January 2014 to 2026. It's amazing how eerily similar the shape is of the blue line—the most recent experience—and the red line—the experience from the pre-Volcker days into the Volcker days. So at least the blue line has curled over, but it'll be interesting to see if we continue to resume something resembling the shape of that prior inflation disaster.
Everyone knows that my favorite inflation index is the Import/Export Price Index because it doesn't have adjustments—it's not seasonally adjusted, it's just prices. And so what we have here is the export prices are now running year-over-year at 8.25%, and the import prices are running at 5.95%. Both are very elevated. If you sum them together, you get something up at around 14%; if you average them, you get about 7%. So based upon these, the purest of the inflation measures, inflation is actually running at about 7%. No wonder consumer sentiment is at such a depressed level.
Here's the Bloomberg Commodity Index, which had a correction in the middle of the year and bounced all the way back, bounced right off that 200-day moving average—the red line—and appears to be in the process of taking out the high of the past 10 years or so, even longer than 10 years. More on inflation: here's the retail prices for electricity to residential customers. I'm not really focusing on the year-over-year number; I'm just looking at the dark line, cents per kilowatt-hour, which was about eight years ago down at 12.5 cents and now it's up to 18 cents, so up 50%. It appears that that line is not slowing down in terms of its trajectory. And so this is another reason why consumer sentiment is suffering and why the poll numbers for incumbents are not what they would like them to be.
Another inflationary issue, of course, is oil, with Brent oil—the true global benchmark—being almost $100 a barrel now. We see the Strategic Petroleum Reserve has been taken way down since the war started. We're now at the lowest level going back to the '80s when this thing started, I guess. We see that we're down to 287 million barrels; we were way up at 750. So we're down by more than 50%. When they refill the Strategic Petroleum Reserve, which should happen at some point, that's going to keep a floor under the price of oil and keep inflation stickier than the Fed would probably like to see. But it's not just the US Petroleum Reserve—take a look at global oil inventories going back to 2018, so going back about 10 years. We see that the dashed line drawing a horizontal line represents the latest level, and it basically is the lowest it's ever been, as low as 2025. So further pressure of keeping a floor under oil prices.
This is a fascinating chart. This is cross-asset performance since the war started at the end of February of this year. What we see is pretty remarkable: there's been fantastic returns from commodities, particularly the energy sector, but the Bloomberg Commodity Index over on the right is up 34% since the war started. We see equities are doing quite well, in particular emerging market equities, and we see Japan is doing quite well. Everything's doing well—everything's double digit. The worst performer appears to be MSCI Europe and the UK, but pretty much everything is up in the double digits and into the 20%. And then you see all the commodities are up, and the Bloomberg Commodity Index, as I said, is up 34%. Then you've got poor old bonds in the middle, with just about the best-performing bond sector being leveraged loans, up 3.1%. Then you've got emerging market sovereign that's up a little bit, and you've got the investment-grade categories—Treasuries, mortgage-backed securities, and corporates—all negative, with mortgage-backed securities being the least negative. So this is very strange. We've got this big donut hole: great returns on the outside and de minimis returns from the fixed income sector.
Debt growth has obviously been a problem. We have the nominal GDP of the United States is the blue line, and we've got all public debt of the United States Treasury. We see that the red line is increasing much faster than the blue line, particularly since the Global Financial Crisis is when it really started to accelerate, and there's no end in sight. The trajectory here just keeps getting steeper. So $40 trillion now in the total debt, including the stuff owned by the Fed and the Social Security system—$40 trillion. The way we're going, we'll probably be at $50 trillion certainly by 2032. Even the Social Security administrators say Social Security, under the current system of funding and benefits, will be out of money by 2032. Of course, they're always overly optimistic on their assumptions, which means we're probably talking about 2029 or 2030. Social Security will have to be reformed, or else cut all of their payments by something like 22%, which wouldn't go over very well, particularly with what's left of the baby boomers at that time who paid in all those years. But we'll see what happens here. We have a really big problem, and of course, it's not getting any better.
This is the annual federal deficits by fiscal year going back to 2021, and we've set a new record here. For a minute there in the middle of this year, we were slightly lower on a year-to-date basis than fiscal 2025, but no, we've set a new record. The pace car here is fiscal 2026, which is ending pretty soon, which is going to start fiscal '27 fairly quickly. It looks like that we're going to set a new record for this fiscal year.
Here's the federal budget as a percentage of GDP, and this is projected out by the Congressional Budget Office. This is projected out to 2035, but the yellow line is the interest expense, and to the right of that horizontal gray line is the future predictions, and they're not pretty. This assumes pretty good assumptions: it assumes interest rates lower than where they are today, budget deficits that are smaller as a percentage of GDP than what they are today, and constant positive real GDP for all this time period. So you challenge any of those assumptions and stress them a little bit, and it's pretty clear that we're headed under the current trajectory to be more like 7% or 8% deficits as a percent of GDP less than 10 years from now. Not good.
Gold looks a lot like commodities. Gold had a huge run into the first quarter of 2026, and then a pretty monstrous correction all the way down below 4,000, and now it's moving back up again. I think gold should be part of every portfolio, and it's quite clear that as the dollar weakens, central banks and institutional investors broadly are preferring gold to fiat currency.
Here's the Shiller P/E ratio, which is at 42, which was higher in 1999, but not by much. You can see that we've got this plotted back to the 1870 period, and this is way higher than it was at the bubble days of 1929. So stocks are not cheap. This is an interesting study: this is a scatter plot going back to 1965 to 2015 of 10-year forward returns based upon the CAPE ratio—the forward real returns based on CAPE ratio. We've got this regression line, that dotted line that's sloping down, and you can say that when you're at this type of P/E ratio, which is now at 42 on the CAPE ratio, you've never had positive real returns in the 10 years forward. In fact, they're significantly negative real rates of return, negative 5% to 9% per year. So we're talking about massive real losses in stocks purchased broadly in the capitalization-weighted S&P at this type of CAPE ratio. It's interesting just how weird that there's so many big negative real returns back there at fairly low P/E ratios going in, but we're more commonly now, in the last 15 years, 20 years, living at more normally higher P/Es than we used to in the past. But this is anything but a ringing endorsement for heavy capitalization-weighted equity portfolios. In fact, I recommend none of that.
Interestingly, the stock market is very concentrated. We all know that with the growth of the tech and AI sectors. Here we have a light blue line, which is the weight of the information technology sector in the S&P 500, and then all of a sudden the light blue line disappears. The dark blue line is what the largest sector weight was. What this means is ever since 2008, tech has been the largest sector in the S&P 500, and it's now at 38% concentration, which is higher than the concentration of the highest sector back in 1999 and substantially higher than prior to the Global Financial Crisis. So there are about as many bargains in the S&P 500 cap-weighted index as there are in a hotel minibar. Here we see the stock market concentrations historically. We've taken this back to the railroads—I'm not sure how good that data is in 1840 or so—but if we look at the 1920s, the Nifty Fifty in the early '70s, the stock market bubble up there in 1987, 1999 prior to the dot-com bust, and then we see what's going on now with the AI Big 10. This is an extremely concentrated market, which means an extremely dangerous market. So I'm not going to recommend any capitalization-weighted equities.
Something's changed also in the inner workings of the stock market. Here we have the rolling 120-day correlation of returns between the AI complex and the S&P 500 ex the AI complex. They were quite correlated from 2021 into 2025 and even in the first half of 2026, but in the last several months this has really changed. So this is about a six-month sort of average if you use 20 trading days per month, and so now we're negatively correlated. As the AI market has done well, the rest of the market is incrementally doing the opposite. They're now negatively correlated slightly. It looks to me like this descent from about a 0.5 correlation earlier this year now down to a -0.14 has got a lot of momentum going, so I wouldn't expect this to reverse anytime soon.
What's interesting is that the equal-weighted S&P 500 has been outperforming. This goes back to 2017, but it's been outperforming for about a year, maybe a year and a quarter. Once something starts to potentially reverse—look back in 2020, an interesting comparison point—we see that we started to go sideways on the relative of cap weight versus equal weight, and then we saw a big correction where equal weight outperformed. Equal weight has begun to outperform. It's not terribly convincing that it's the start of a big trend, but it's certainly not underperforming any longer.
Also, US stocks versus the rest of the world are no longer outperforming. Take a look at US equities versus non-US equities: when this line's going up, it means the US is outperforming, and when this line is going down, it means ex-US equities are outperforming. It's kind of sideways for the past year, year and a half, but it clearly has stopped outperforming. We'll zoom in on that: this is the same chart, but we're using a shorter time series. We can see this actually started two years ago. The US equity prices had their maximum outperformance almost two full years ago now, and had a pretty big underperformance into the middle of 2025 and the first quarter of 2026. But I think that this line on a trend basis is headed lower, so I think that makes sense in a longer-term perspective, not just a short-term perspective, to think about foreign equities. I had been investing in foreign equities. Right now, I'm going to come close to home because I don't like the way the risk setup is for the markets.
The dollar has been falling ever since the end of 2024 when the dollar was at 110 on the DXY index. It fell down below 100 where it resides today. It's been eerily stable. It almost looks manipulated—I mean, it hasn't changed any meaningful amount in over a year. But interestingly, with the dollar falling, we see that non-US stocks have started to outperform. Global price-to-book ratios are way out of whack. This is a valuation argument against US equities: the MSCI US Index has a price-to-book ratio of 5.72; the rest of the world excluding the US—so everything else—has a price-to-book ratio of 2.49. You might believe that the US is the best thing ever in the history of the world, but you'll notice that there have been time periods during corrections in particular where the tan line and the light blue line converge, which would lead to a massive underperformance of the MSCI US Index versus the MSCI World Index ex-United States.
Here's the S&P 500 versus MSCI only the emerging markets. Here the underperformance is pretty obvious: the outperformance of the US stopped at the end of 2024, and it's underperformed by about 20% now. So it's not insignificant, and I further believe that this is going to drop going forward. Here's the S&P 500 relative performance to the MSCI EM Index—that's the red line. When the red line's going up, the S&P is outperforming the emerging market index, and when the red line's dropping, emerging markets are outperforming. The blue line is the US Fed trade-weighted nominal broad dollar index. You can see that the red line and the blue line are very similar in shape. So if the blue line falls, which would be the dollar trade-weighted nominal broad index dropping, it's quite likely the S&P 500 will underperform emerging markets.
I'm also sensitive to the seasonals right now. Here we are early September, and September and October are generally difficult months for risk assets. So that's one of the ideas that's informing the recommendations I'm about to go into. Here's US local debt—this is bond market now versus US corporate total returns relative to the JPMorgan Emerging Market Local Currency Index. The tan line is the local currency index versus Bloomberg total returns, and we've got the dark line, which is the dollar index inverted. So when the blue line is going up, it means the dollar is going down. Again, the tan line and the blue line are very similar in shape. So if the dollar goes down, which I expect to happen, we'd expect EM local currency emerging markets to outperform US corporate bonds. So with that, here we go barreling into the holiday season. NFL starts tomorrow. Go Bills! Thanks everybody for your participation in this call. Thank you for your support of DoubleLine, and goodbye for now.
