news

한어Русский языкEnglishFrançaisIndonesianSanskrit日本語DeutschPortuguêsΕλληνικάespañolItalianoSuomalainenLatina

On July 13, in the monthly series "In the Know", well-known investor Cathie Wood, Tasha Keeney, director of investment analysis and institutional strategy of her Ark Investments (ARK), and Sam Korus, director of autonomous technology and robotics research of the company, discussed how ARK values ​​companies and makes investment decisions. Cathie Wood also expressed her views on issues such as U.S. Treasury yields and U.S. stock concentration.

This article collects the wonderful views of the three people in the show:

1. ARK adheres to a combination of top-down and bottom-up evaluation methods and conducts valuation based on business lines.

2. In ARK's scoring system of 1 to 10 for companies, if one score is lower than 7, it is worth paying attention to; but if two scores drop, we will seriously consider whether the risk of the portfolio should be reduced.

3、Most of Tesla's commercial value comes from Robotaxi.The latter will change Tesla's business model: from the current one-time revenue from selling electric vehicles to recurring revenue from RoboTaxi.

4、We think Tesla could be the first automaker to roll out robotaxis on a large scale.At scale, robotaxis could cost as little as 25 cents per mile.

5、Tesla's ups and downs depend on its executionThe key factors that may have an impact include: key personnel (Musk), personnel changes, moats, cutting-edge technology impact/adoption, battery costs, etc.

6. Due to the cash generated by autonomous driving technology and its successful production scale, Tesla is expected to expand its annual car sales from less than 2 million vehicles currently to 6 million to 14 million vehicles.

7. We believe that Tesla is very likely to launch a robot taxi network in the next few years, at least in the next three to three years.

8. It is expected that by 2029, the valuation of Tesla's electric vehicle business to EBITDA ratio will shrink from the current 60 times to 19 times, and the growth rate of revenue growth and profit expansion will remain strong.

9、In the long term, robotics, energy storage, artificial intelligence, blockchain technology and multi-micron sequencing technology will increase real GDP growth from 3% to6% range.

The following is a transcript of the conversation:

Cathie Wood:Hey everyone, today we're doing something a little bit out of the norm - we hope to be launching a new segment each month that shines a light on some of the stocks and companies we're focusing on in the innovation space.
Today, we're going to be specifically talking to you about how we do our bottom-up analysis. I think we're known for analyzing the market from the top down using rights law. But we wanted to explain a little bit more about our analysis, not just from the top down, but also from the bottom up, and specifically our valuation scoring system. To that end, I'd like to introduce Tasha Keeney and Sam Korus, who do great work at Tesla. Tasha is our director of investment analysis and institutional strategy, and Sam is our director of autonomous technology and robotics research.
So we wanted to walk you through our valuation system, and we're going to use Tesla as an example because we've published a lot about it. We haven't published much about other stocks yet, but we're going to talk about Nvidia in another webinar. Next Thursday is our quarterly webinar, so stay tuned. After that, I'll be doing monthly non-commercial case presentations. Now, Tasha, Sam, and I are going to walk you through the valuation system. Tasha?
Tasha Keeney：Thanks, Cathie. As you all know, we at ARK are very focused on disruptive innovation, and our investment team is allocated by technology. So Sam and I have been working together for almost a decade and are now looking at autonomous driving technology and electric vehicles. We take into account all the top-down analysis we do in both areas, and we'll talk about that later.
And then the bottom-up work that we do once we find potential names for the portfolio is.Bottom-up work includes both qualitative and quantitative indicators based on assessments.On a quantitative basis, you're well aware that we've published the Tesla blog with our 2029 price target. Again, that's a combination of top-down and bottom-up. So, Sam, why don't you first walk us through the top-down research that you've done on electric vehicles?
Sam Korus：Absolutely. And this ties back to what Cathie was saying earlier, really leveraging the empowerment law and the cost agency. So look at the battery cost declines and illustrate what we think is going to happen to electric vehicles in terms of price parity. It's already happening in China, and you can see their adoption rates are way ahead of the U.S., where that transition hasn't happened in many areas. All of that leads us to our top-down forecast, where we expect electric vehicles to expand from about 10 million units in 2023 to 74 million units. So by 2030, probably nearly all of the automotive market share. And then we combine that bottoms-up, and that's how we get our differentiated outlook.
Tasha Keeney：In terms of the prospects for self-driving cars,We think the robotaxi industry could be worth $28 trillion by the end of the century.That's why when you look at our Tesla valuation,We think most of Tesla’s value will come from this as it will revolutionize the business model- Transition from the current one-time revenue from selling electric vehicles to a recurring revenue stream from Robotaxi.
Of course, we believe robotaxis are disruptive because they could be priced lower than all of today’s ride-hailing options and far lower than the cost of driving a private car in the U.S. today (about 70 cents per mile). So at scale, robotaxis could be priced as low as 25 cents per mile.
But initially, our price target was higher. So when we evaluate companies like Tesla, in our evaluation, we look for every company in our valuation to be able to achieve at least a 15% compound annual growth rate over a five-year period. When we do our initial evaluation of Tesla, we evaluate it qualitatively. So you can see on the slide that we first look at people management and corporate culture. That factors into our thinking about the management team. Sam, why not start with Tesla?
Sam Korus：Yeah, I would just say that these qualitative scores are often backed up by quantitative measures, so you'll see a lot of productive discussions among investment teams about this.
So for people, for management, for culture, you know the way we look at this is we look at it through the lens of disruptive innovation. So we actually want a founder who is also the CEO who is willing to make long-term bets and potentially invest in the short term, which may not be what the market wants or what people who are more short-term oriented want.
But for Tesla, having someone like Elon Musk who has a big vision, who has master plan after master plan, and who is willing to sacrifice short-term profits to pursue huge opportunities like autonomous driving, we think that is a huge positive.. So while it's great that Elon is a visionary founder, we also look at departures and our ability to acquire talent. So, you know, Cathie, you're a key member of the Tesla team who review these conversations that we've had.
Cathie Wood:Yes, the departure of key people was asked a lot. For example, after JB Straubel (note: Tesla co-founder and long-time CTO) left, who was really responsible for Tesla's battery and drivetrain strategy? He left the company, and we received a lot of similar questions. Someone very important to the whole story left. What's interesting about this departure is that JB left after his main work was completed, which means that the battery part and the drivetrain part of the story were in place, and now the main driver of the strategic vision will be autonomous driving. Andrej Karpathy (note: Tesla AI director) worked at Tesla, then parted ways, and now he is back. In short, it is a very difficult environment and many people are exhausted.
So we do see people come and go, and we understand that. But Elon is clearly the visionary. He's surrounded himself with some of the best engineers in the world who want to work on very difficult projects, the most difficult projects in the world, which is the AI ​​part of the strategy. As you've heard,We describe Tesla as the largest AI project on the planet, we believe it is attracting the best talent in the world.
Sam Korus：Yes, in fact, in an annual survey of the companies that recent graduates most want to work for, SpaceX and Tesla still rank at the top, which is a good proof of this.
Tasha Keeney：We've talked about this in the past, and the idea is actually that having Elon at the helm of many companies actually does help with talent acquisition. So, as Cathie mentioned, while it's a difficult place to work, it's also a very desirable place to work. That's why, you know, there are rumors that Andrej Karpathy might come back.
Cathie Wood:I want to add one other thing here. You'll find that from a risk control perspective, you can see what impacts our scores.
I have just given an example of the resignation of key personnel.JB opened the door, so to speak, and his departure comes at a time when the electric vehicle story is accelerating. But another area of ​​risk is around governance, and specifically at Tesla. Of course, we evaluate governance at any company. But to give you an idea of ​​what we're up against here, you'll remember when Elon tweeted at the time that he was considering taking the company private. He tweeted that the funding was secured. Of course, that drew the ire of the SEC. In fact, there were some lawsuits that occurred. So when that happened, we learned that the SEC was indeed investigating whether the funding was secured, and if that was an accurate description of what was going on in the organization, we knew that it would be a distraction from the company, and we did lower the score for that reason.
So we do adjust the scores when we feel like something poses a risk, but what was interesting at the time was that a few weeks after we moved that score down, Tesla released the specs for its own AI chip, and we looked at them and said, "Oh my goodness, this is much more advanced than we expected and it's coming sooner than we expected," and so the other score, "Moat," was moved up again.
So you can see the dynamics of our scoring system. What I'm saying is,On this 1 to 10 rating system, anything below a 7 is cause for concern.The governance level at that time did not drop below 7 points, and the "moat" score increased.But if both scores drop, we would seriously consider reducing the risk of the portfolio.So Tasha?
Tasha Keeney：Thanks, Cathie. Yeah, so I'm going to turn to "Moat." You know, some of the things that we've thought about here over the years are the vertical integration advantages of Tesla, we think they're years ahead of their peers in batteries, AI, data, and as Cathie mentioned, Tesla also produces hardware, chips, and inference chips in-house. Sam, talk to us about battery integration at Tesla.
Sam Korus：Well, I would say that, especially when a company is at the forefront of a new revolution, there may not be a supply chain. So Tesla's investment in battery manufacturing chemistry, and through power inverters, really differentiates its performance from its peers. The current situation remains thatOnly one or two other competitors have similar efficiencies to Tesla, but none have been able to vertically integrate as quickly as Tesla has.
Tasha Keeney：When it comes to autonomous driving, Tesla has access to billions of miles of driving data, while Waymo, as a competitor, has only millions of miles of data.So this is an order of magnitude difference.。
This data will be used to train the neural networks that will allow Tesla to layer in autonomous driving features that we believe will ultimately create a fully self-driving car where you can actually take the wheel off the steering wheel. This is, of course, the basis for a robotaxi network.
Another point I want to make is thatThe Supercharger network also provides Tesla with a great moat.Of course, it's now the North American charging standard, and many other automakers have signed up to be part of the network.
Cathie Wood:Just look at the risks of the rise of new disruptive technologies. I can't tell you how many times we've been in a situation where, oh my god, there's a new type of battery, and Tesla is going to be left behind because it's locked into its current strategy. We know how difficult battery technology is, and the odds are very high that Tesla will adopt new battery technology because its expansion is built on batteries, so of course it wants to stay ahead.So we certainly consider the risks of new technologies emerging.We certainly talk to Tesla about emerging technologies, but every time we are assured and reassured that they are very knowledgeable about what is happening in terms of new technologies, particularly on the battery side.
Tasha Keeney：Yes. Another important qualitative score that we judge every company on is execution. Sam, tell us about production hell. Give us a little history here.
Sam Korus: I think Tesla's ups and downs depend on its execution.You know, during the ramp of the Model 3, Tesla was in what was called "production hell." I think we all remember that time well, and when we look at this bright future, they missed X thousand deliveries, the stock price plummets, and the thought is, "Well, they missed a few thousand deliveries this quarter," and the question is, is this quarter significant? Or is there something wrong with the long-term story? Are they still expanding?
I think with Tesla, we're seeing a real execution momentum where they've been through tough times and now even Toyota is saying, you know, they make one of the best cars in the world. That's high praise from a company that's known as one of the best manufacturers in the world.With China's help, Tesla has been able to expand its factories faster than anyone else in the world.
therefore,I think on the electric vehicle side, we really saw continued strong improvement in execution.But again, this is something we talk about often, good times and bad times. Maybe, Tasha, I can throw this question to you because one place where there's definitely going to be starts and stops is autonomous driving and the claim that it's going to be here this year, what are we seeing in terms of execution on autonomous driving?
Tasha Keeney：Tesla's plan to launch a robotaxi network is really a quest for the next big milestone in our opinion. It's not an easy thing to do, right? It's very difficult to build a fully self-driving car. We know it's possible. Because competitors like Waymo have already launched it on a small scale,But we think Tesla could be the first automaker to launch it on a large scale.Tesla has already released some of the most advanced driver assistance features to date, and similarly, we think they can stack these features to create a fully self-driving car.
Going back to Sam's point about how qualitative and quantitative interact, when we talk about all of these execution metrics, that's what we factor into when we model Tesla on a valuation basis. We think that because of the cash generated by self-driving technology and its successful production scale,Tesla could scale from less than 2 million cars sold per year today to, in our bear case, about 6 million. In our big case, perhaps 14 million cars.So when we model the company, again, we take all of these metrics into account.
Sam Korus：I also want to say that,Another area of ​​execution we saw with Optimus is leveraging both autonomous and manufacturing technologies, right? It's a real convergence. I think the speed with which they've executed it has surprised everybody. You know, if you go back a few years ago, people would have laughed at the idea of ​​someone dancing on stage in a robot suit. Now, Tesla has at least a few robots working completely autonomously in the factory. It's been amazing to see this come to fruition.
Cathie Wood:I want to emphasize a little bit here about execution, especially as it relates to Tesla. We know that Elon is very big on his vision, and a lot of times he'll present his vision in "Elon time," and I think we all have to adapt to Elon time. But why does he do that? He does this not only to signal to suppliers, "Hey, this is my work schedule." Get all the suppliers ready, but also to motivate employees and keep them focused on achieving their goals in a very short period of time. So,Another thing to watch as we execute, you'll see the first point we focus on is the lack or decline in R&D.
Now we look at this from two sides. R&D is 1% of sales, where are they spending that money? Because sometimes we disagree on a direction. I know when Toyota was moving from electric to hydrogen fuel cells, we owned the stock at the time and saw that they were moving their R&D money to hydrogen fuel cells. We had looked at how much more expensive the infrastructure for hydrogen fuel cells was going to be, especially compared to electric fuel cells. So we sold the stock based on that.
Tasha Keeney：Great. You know we've touched on some of these points, but let's move on to product leadership. I think in the electric vehicle industry, we have a lot of evidence of Tesla.
Sam Korus：Yeah, again, it's an ongoing debate and a quantitative debate. You know, Tesla's market share in the U.S. in different regions is certainly questionable as other automakers continue to launch their products. But in the comparison to Apple's profitability, we can see that as the industry develops, Apple's sales are declining, but profits are increasing. So these are ongoing conversations and debates internally. People also like to discuss these issues on Twitter. In any region where Tesla's stock price is down, you can see people discussing this issue. So this is something we watch closely.
Cathie Wood:One of the things that has happened recently is that GM and Ford have pulled out of electric cars because they don't see profitability. Ford is known to lose $100,000 per car, and maybe Tesla's share in the U.S. will be higher than we think in the next few years. Of course, it's going to come down from 90% or 95% everywhere. Just like Apple did with smartphones.
But the signals we're getting from other automakers about how committed they are to this space is very interesting to see the pullback. The only way to be profitable is to have scale. So it's a very interesting time for us from a market share perspective.
Sam Korus：And then, Tasha, you've touched on it, and maybe also on the autonomous side, I know just this morning we had a big discussion comparing Waymo and Tesla and the lead in autonomous driving products.
Tasha Keeney：Yeah, you know, to tie into Cathie's point, when we look at the traditional automakers like GM and Ford, they've pulled back on autonomous driving - Ford sold its autonomous driving unit, GM's autonomous driving unit Cruise temporarily stopped operations - even though now they're trying to restart.
So when we look at the competitive environment in the U.S.,Really just Waymo and potentially Tesla.Again, Waymo has hundreds of cars on the road. Tesla has millions of cars on the road that they believe can be turned into self-driving cars. So, as I mentioned, they already have greater scale in terms of data advantage, and we think they can turn that into an AI advantage on the road, and they can launch a robotaxi network that will be in many cities, not just the select cities that Waymo has been able to launch service in so far.
The last point I want to touch on here is theoretical risk. Again, we discussed some of the key points in this regard. In terms of valuation of theoretical risk, we did a Monte Carlo analysis (assessment of the possible risk impact in various real-world scenarios) for Tesla. We have a lot of variables and we change the upper and lower limits of these ranges based on the probabilities we think are appropriate. For example, we think it is very likely that Tesla will launch a robot taxi network in the next few years, at least in the next three, two, three years.
We also factor in the timeline slippages that we talked about, right? Elon has promised multiple times that they're going to build a fully self-driving car, and that hasn't happened yet. And again, crossing that threshold is a very difficult thing to do. So we can't be absolutely certain that the launch of the robotaxi network is going to be July 9 or any specific date. But we try to use our research and our analysis to come up with a time frame that we think is likely to happen.
Cathie Wood:I want to add something about regulatory risk. When we first started the Tesla journey, we thought regulatory risk was very high. But what's happened in the last decade since we started this is that the crash rate in the U.S. has gone up dramatically after decades of decline. I think the lowest number of car fatalities in the U.S. was about 30,000, which was about 2014. Since then, it's gone to 45,000. That's alarmed regulators. So now they're looking at the data. And Tesla is able to provide a lot of safety data.
I'll hand this off to Tasha, who has done a lot of research on safety. It's very interesting that this doesn't get more attention because Volvo's entire brand is based on safety. Some of the numbers that Tasha found in her research are pretty staggering.
Tasha Keeney：Yes, again, this is something that we can quantify in our analysis. We estimated a long time ago that self-driving cars could be 80 times safer than today's manually driven cars.In terms of accident rate, safety is 80% higherWe actually have data to prove it. So we've seen that Waymo is actually safer than the national average. And when you look at Tesla,A Tesla driven by FSD is approximately 5 times safer than a Tesla driven by manual.This is actually based on the latest available data points.
Elon has said that they think sight lines are probably going to be four to five times better. When you compare that to the national average, it's about 14 times safer. So, you know, we've seen evidence that autonomous vehicles are safer than human-driven vehicles. And we think that's what regulators are looking for when they decide whether to allow these vehicles on public roads.
Another point I'll touch on is that Tesla is in discussions with China about the FSD rollout, although nothing has been decided. But if that's true, it would be a major step forward. Sam, why don't you talk about key person risk? That's another important component of theoretical risk.
Sam Korus：We've certainly discussed Elon's central role, particularly in terms of fully leveraging the vision of driving autonomous driving. So we do think about that and the importance of his role in launching Robotaxi, and I think even he's said that there comes a point where it doesn't make sense for him to be the CEO and take on a more visionary or product-driven role. But I think we've discussed that,It is important to drive autonomous teams towards commercial launch.
Cathie Wood:So, Tasha, let's wrap up the valuation question and bring the valuation exercise we did into reality.
Let me set the stage. The metric we chose is enterprise value to adjusted earnings before interest, taxes, depreciation and amortization (EBITDA), and we adjust EBITDA in two ways:
One is equity incentives. We do think it's very important to incentivize employees, including Elon, by the way, and his compensation package, and to really align them with shareholders. So the second adjustment is on the R&D side, which is to normalize the R&D standard. Our companies typically spend too much on R&D, and of course we expect that. We expect them to sacrifice short-term profitability because especially in the AI ​​world, the winner takes all.
Therefore, our hypothesis is thatOver the next five years, EV business valuations will deteriorate as a ratio to EBITDA deteriorates toward market multiples, and our analysts believe that a combination of revenue growth and profit expansion will outweigh the valuation decline.Tasha, can you talk a little more about this?
Tasha Keeney：If you look at Tesla's electric vehicle business, the EBITDA today is about $57 and it's closer to $60 in the final year of our model. We have our own valuation multiples for each business line, and by 2029, the largest piece of the value pie that is assessed under this business line is autonomous driving, which is expected to be valued at 19 times EBITDA by that time. So, as Cathie mentioned,It will be greatly compressed from the current 60 times to 19 times.
In fact, in other business segments including electric vehicles, we assume lower multiples, about 12 to 13 times. So the blended average multiple is even slightly below 19. So based on how we value the Tesla business again, it can basically be identified as a mature business, and in fact,We believe Tesla will continue to maintain a respectable growth rate in the fifth year of this model.Therefore, our valuation of Tesla in five years is actually a relatively conservative assumption.
Cathie Wood:Again, we set a minimum annualized return of 15% over a five-year period. So when I say 60% to 18%, our revenue growth and margin assumptions have to override that. As you all know, in our empowerment work approach, we are very rigorous about how to model long-term cost reduction and scale expansion. So I think we will discuss it here, perhaps in an abbreviated version of the Global Report. No, we will not spend much time on fiscal or monetary policy this time. Fiscal policy, for obvious reasons, is in a state of confusion in Washington.
Chairman Powell did seem to emphasize monetary policy in his latest testimony, and he's starting to focus on employment now. He's looking at some of the statistics that have been released and said that he also acknowledges that inflation is continuing to ease. This week, Austin Ghoulsby, who I think is probably the biggest dove, said after the CPI report, "Ahaha, this is what I've been waiting for." So we'll just talk about economic indicators, last week's employment report, the headline was again stronger than I think the consensus expectation. But I think the consensus expectation was 190,000. It turned out to be 206,000. But when adjusting for another downward revision, the actual number was 95,000, and household employment increased by 116,000 after a decline of 408,000 in the previous month. So household employment was weak, flat to down year-over-year, while nonfarm payrolls were still just over half a percent. The decline in temporary help accelerated, down 49,000. That's a pretty big number for temporary help. So I think Chairman Powell took note of those appropriately.
Of course, the other variable that the Fed watches is inflation. We got CPI this week, and it was below expectations, which doesn't surprise us. As I mentioned in my last post, no, we're seeing price cuts, not just lower inflation, but price cuts, Walmart Costco, Best Buy, McDonald's, Wendy's, even Starbucks had promotional packages, which is very unusual. So CPI actually fell 0.1%, which was the headline, bringing the number to 3% year-over-year, not yet at 2%, but we think it's going to get there. Core was up just a little bit. Yes, 0.1%, which was also a little below expectations after the 0.2% increase last month, and the core is now up 3.3% year-over-year. Again, we believe this will get to 2%, and I mean close to 2%. We also got PPI, which was a little hotter than expected.
What does this tell us?This tells us that margins are compressed.So for many observers, they're concerned that we're going to see cost-push inflation, like we saw in the '70s, where costs keep going up, they're forced to raise pricing, and consumers don't accept the higher prices. We're seeing some companies, even PepsiCo, which just this week announced that its North American sales were down 4%, which is very unusual for a consumer staples company.
Now we do have some unusual things going on, COVID is probably one reason, but we do think the economy is another reason. So it's probably a 50/50 split, maybe, who knows. We think Pepsi is going to have to respond with lower prices.
Now I'm not going to list all the economic statistics. What I will tell you is that when I look at it and prepare for this, I think almost all of the statistics are weaker than expected, especially housing and capital spending. Both of these have high multipliers associated with them.
Interestingly, one stronger-than-expected number that came out right now was personal income up 0.5%, but spending up just 2%. What does that tell us? It tells us that the consumer saving rate is starting to rise. When does that happen? It usually happens when people are generally concerned about their jobs and the economy.
So I'm going to show you a chart about this in this set of charts. So we'll start with the chart right now. Okay, you can see here what the market is struggling with.
This is a long term chart of the 10-year Treasury yield going back to the early 1960s. You can see that the trend from the 1960s to the early 1980s was consistently up and moving significantly from the low to mid single digits to over 15%.
Since then, we have been in a downward trend. You can see that the downward trend seems to have ended. And this is the reason why there is a lot of controversy. One of the reasons why it ended is that the Fed raised interest rates 22 times to fight inflation, and it decided to be hawkish because it felt that it was lagging. This is one of the reasons why long-term Treasury yields have risen. If it had not risen, we would have had an inversion of about 500 basis points or 5 percentage points in the yield curve, which is a very unusual situation.
Now the question is, have we broken this downward trend? The answer is probably yes, because we are close to zero, and negative interest rates are not a good thing. This may be related to a very difficult economy.We expect the economy to be weaker than most expect in the short term, likely due to consumer spending.The economy type pricing we were expecting was much lower than expected in the long run.
In fact, we are seeing a trend towards deep deflation brewing in the short term for cyclical reasons, pricing has gone too far due to supply chain shocks. Consumer product companies have taken advantage of this, and now they will have to lower prices, which is cyclical. In the long run, there is a long-term reason for deflation, which is innovation,True disruptive innovation is deflationary in nature.
It follows a learning curve. We see increasing efficiency and productivity gains, and they show up as the cost of the technology itself goes down, leading to a lot of what are now mass market opportunities. So we're going to see prices of these technologies go down, but unit volume growth goes up significantly. So nominal GDP is probably going to be in the 5% to 10% range. Over time, long term, Treasury yields have been aligned with nominal GDP growth. So we think in the short term, for cyclical reasons, Treasury yields will be lower.
Yes, in the long term, we do believe that the long term downward trend in U.S. debt has been broken because we believe the unit growth from these new technologies will be substantial and could drive real GDP growth to surprising levels. In past webinars, we have shared that we believe that over the next 10-15 years, as multiple innovation platforms develop simultaneously, real GDP growth could be higher than the 3% range that we have seen over the past 125 years since the last period. In the early 1900s, with the advent of the telephone, electricity, power, and the internal combustion engine, GDP growth went from 0.6% to 3%. We believe thatIn the long term, robotics, energy storage, artificial intelligence, blockchain technology and multi-micron sequencing technology will increase real GDP growth from 3% to6% range.
The next chart deals with the cyclical part of the economy. We are watching the ratio of metals to gold very carefully. You can see how close the correlation has been since 0.8. When the price ratio of metals to gold goes down, they mirror each other, and until recently, U.S. Treasury yields also went down. And then you can see recently that correlation has been broken. The ratio of metals to target prices has gone down. It's gone down by 0.8, 0.9 levels. But yields haven't gone down yet, and we think they will.
As I mentioned before, for cyclical reasons, we were waiting for some signal from the metals to gold ratio to lead us in another direction. That did not happen. So, we do believe that the 10-year Treasury yield will fall. I want to remind everyone that we are in a long-term downtrend, and the actual downtrend started after energy prices, $0.8, and oil prices soared to the 47 price range. Since then, it has been a downward trend. The COVID-19 pandemic has brought it to very low levels. The supply chain shock has increased again, and it is now in a downward trend. Commodity prices today are at the same level as in the early 80s. This does make some sense.
The next chart is something I've already described, but I'm describing it last here, which is the yield curve. So this is measured by the 10-year Treasury yield relative to the 2-year Treasury yield, and you can see that the 10-year Treasury yield is below the 2-year Treasury yield. This typically precedes a recession. The shaded area on this chart is a recession. You can see that when we see an inversion of the yield curve, we typically have a recession within a year and a half. That hasn't happened yet, although I do think that with all the downward revisions we've seen in the employment indicators and many other indicators, the NBER's rating of a recession could be backdated to sometime late last year.
So the other thing you can see is the downward trend in the yield curve since the GFC. I think during Covid, you can see that we did see a steepening of the yield curve, meaning it went from about 0% to 1% a/2%. I thought it would go higher like it did during the GFC. Why didn't it? Every government around the world, both monetary and fiscal policymakers, was doing their best to tighten financial conditions, but it should have been back in that 200 to 300 basis point range. Why didn't it? We can see clear deflation in commercial real estate, in the office space, and we're increasingly thinking about the multifamily space. So that's one of the sources of cyclical deflation.
I mentioned the other issue earlier, the consumer, pro-consumer companies cutting prices. Here we are upside down, and you can see that we have been upside down for a long time. In fact, it took longer than any time since Fed Chairman Volcker in the late 70s, early 80s to try to curb inflation that had reached double digits. We never reached double digits, but Fed policy on this basis looks as tight as it did then. That's why I keep coming back to the reason that there may be downside risks and the reason that the Fed is going to have to change direction.
The Fed has four more opportunities to reverse policy this year. Later this month, in July, September, November, and after the December elections. We wouldn’t be surprised if three rate cuts occurred as these deflationary forces build over the next few months.
On the next page, I want to show the adjusted corporate tax rate, which is from the GDP accounts, and you can see that margins have really been growing since the early 90s and now seem to have peaked. We think margins are going to fall more because if companies want to maintain unit sales, they will be forced to cut prices.So we do think that margins could decline to below 2014 levels.
That said, we think overall that the upward trend or at least high levels will continue because of all the new technologies, especially AI, artificial intelligence and the productivity boom that we see in the future. So the margin decline will be more cyclical. But in the long term, we think profitability will remain elevated. Maybe we won't continue to go up. It's already pretty high by historical standards, but all of this new technology will increase margins.
The next page is just to illustrate how low the consumer savings rate is in historical context. You can see that during the COVID period, it went up to about 32% because no one was spending money. Of course, the fiscal stimulus also provided the necessary funds. People were saving a lot at the beginning, but now it has been completely depleted. Our savings rate is on the low end by historical standards. You can see in the shaded area, these are the savings rates that rose during the recession.
We just got another reading on the savings rate. It's up to 3.9 now. I think it was 3.6. So the reason for that, as I mentioned before, is that spending growth is lower than income growth. So that's a caution signal on the next chart.
Now we're going to talk about market indicators. This chart is a chart that Goldman Sachs made, and it's a measure of concentration in the stock market. Maybe you've heard of the tech "Seven Sisters," which has become the "Six Shows." When Tesla started to underperform this year, the people who named it kicked it out. So Mag 6 leads us to a record of market concentration. A lot of people think, "Well, I feel safe with Mag 6," they have great cash flow. They seem to be in the sweet spot of artificial intelligence. But we would say that the more concentrated the market, the higher the risk of these stocks by definition.
Now we looked at this chart at the beginning and said, gosh, when we got to this high concentration, my eyes fell on two points on the chart: 73 and 2000, because I lived through it. Of course, 2000 was the peak of tech dot-com, and then we had the bear market. So this concentration is a precursor to the bear market. Go back to 73. I wasn't involved in this market at the time, but I read about it, and when I got into the industry, it was just folklore. That's the end of the Nifty 50 index. The market was concentrated in 50 stocks again, and the result was a terrible bear market and the worst recession since the Great Depression.
So I look at these and say, if you look at other concentrated peaks, you'll see that they tend to be associated with huge fears of extended bull markets.
Let's illustrate. By 1932, the economy was in recession, the Great Depression had begun in 1929. Of course, unemployment was at 25%, not 4.1% as it is today. Real GDP was down 30%, consumer and producer price inflation was down, real prices were down 25% to 30%. Of course, companies with huge cash cushions and the ability to generate cash flow even in a recession were in the lead and received a disproportionate amount of capital flows. So what happened? We entered a bull market.
I think the market is going to go up more than 50% or 60% over the next three and a half years. And it's going to be significantly more broadly., so much so that small-cap, mid-cap and even large-cap stocks at that time were outperforming by a wide margin. Large-cap stocks have done a lot, and they are treading water. And others that were hit hard in the first three years of the Great Recession have done better. We think that's true now. Of course, this will benefit strategies like ours that have little overlap with the broad benchmarks, and right now Mag 6 is disproportionately represented in our category. All of our stocks are focused on disruptive innovation.Therefore, as interest rates fall, our long-term bonds should be the main beneficiaries of an extended bull market above.
As you know, we believe that inflation will continue to cool more than expected and that the consumer has essentially filled the rolling recession that we've been in since the Fed started raising rates, and the housing market has been down almost randomly and it's still down and it's actually relapsing. We're seeing a slowdown in housing, commercial real estate, autos and now broader consumption.
Here's a chart that gives us an idea of ​​how fast reversion to the mean can happen. So you can see that with the rise of the tech internet, value stocks outperformed for quite a while. And then all of a sudden, the bubble bursts, and in less than a year, value stocks went from underperforming by over 3,000 basis points to outperforming by almost 5,000 basis points. So as we got closer and closer to the end of the bubble, a lot of portfolio strategies shifted their style toward the internet. Those that did that probably lost a lot of business. The strategies that stuck with their style experienced a huge shift, reverted to the mean, and then outperformed. And interestingly, as the dot-com bust happened, we had another reversal, reverting to the mean again.
As you can see here, value stocks have been underperforming recently, as have long-term investing strategies like ours. We believe both strategies will revert to the mean as interest rates fall.
Finally, I'm going to show you one of the Mag 6 stocks, Apple. You can see that what's striking to us is that Apple is valued much more today. I think it's three times. If I'm reading this correctly, that's much higher than when the iPhone was launched, which should have been a big jump in valuation. Instead, what's happened recently is that its cash flow, especially with interest rates going up, looks very attractive. I've never seen the market pay for cash flow, but that's what's happening because Apple's revenue growth is zero. As it stands right now, Apple's revenue growth has been very low, it's been zero for the last several years.
Now, it's getting some boost now through its partnership with OpenAI, or at least opening up to OpenAI and other foundational large language models. We do think there's going to be a refresh cycle. We think AI will add nice features and functionality, but we think this multiple revaluation, just because it has cash, is going to give way over the next few years, especially when interest rates come down, as we believe they will. So the valuation boost that Mag 6 has had over the last few years, is because of the cash. What that tells us is that there's a lot of concern in the market that the reallocation into equities has been disproportionately into safe stocks with higher cash levels. We think as interest rates come down over the next few years, we're going to see a reversion to the mean again.
Now, it's getting some push now through the partnership with OpenAI, or at least opening up to OpenAI and other foundational large language models. We do think there will be a refresh cycle. We think AI will add nice features and capabilities, but we think this multiple re-evaluation, just because it has cash, will give way over the next few years, especially when interest rates come down, as we believe they will.
So the valuation of Mag 6 has increased over the past few years because of cash flow. This tells us that there are a lot of concerns in the market that the reallocation of stocks has disproportionately flowed into safe stocks with higher cash flow levels. We believe that as interest rates fall in the next few years, we will see a reversion to the mean again.