Historically, companies were valued with a heavy weight put on their financials. That doesn't seem to be the case anymore, and without any clear approximation of how we are valuing companies it feels to me more like pure gambling.
What would be an argument that supports this?
To me, it seems fairly easy to imagine that the p/e of the S&P could increase forever. Either because the increase of the money supply is constantly accelerating or because our ability to increase productivity is constantly accelerating.
The p/e ratio is an interesting indicator because it more directly ties together the how much money the company actually makes relative to what I'm paying for it.
P/E ratios over time can be good indicators of relative value of the company. I'm not exactly inventing economic or investment theory here to point to high p/e ratios as an indicator of overvalued companies.
For fast growing companies, it’s useless because they quickly add new revenue streams and profits can change quickly.
If you are purely investing off of sane P/E ratios, you might make some safe investments, but you will always be late to the party.
high p/e ratios as an indicator
of overvalued companies
Imagine an investor in 1988 (the middle of the p/e chart you linked to) who took your approach. From their perspective, the historical average p/e of the S&P 500 is about 15. Would they have played the market by selling over 15 and buying under 15, they would have missed almost all of the nice long rally that came in the decades after.
> they would have missed almost all of the nice long rally that came in the decades after.
I wasn't saying to buy and sell as the ratio goes above/below some specific threshold. More importantly though, this is ultimately cherry picking. You can always look back at historic data and define a start point, or a start and end point, to make an argument work. That has little, if any, weight on predicting what will happen in the future.
I'm being pretty careful here to not claim to know what prices or the economy will look like in the future. I'm simply sharing my read on p/e ratios, what my own take away from they is, and where what I see as historic outliers are.
What do you mean by that?
When I'm investing in a business I want to know how quickly I can get my money back, and how risky I think the investment is relative to my expected gains.
When I'm investing in baseball cards I want to know how much the next person will pay me for it. The stock market seems much more inline with that today, and in any reality where stocks are meant to be considered investment that seems disconnected or entirely unhinged.
Let's talk about a specific company then. Which one have you looked into recently?
By no means am I saying that is predictive, maybe the ratio isn't as useful as I think or maybe it is less applicable today for some reason. I do have a hard time finding any stocks today that I can justify the market valuation for, though.
[1] https://www.currentmarketvaluation.com/models/price-earnings...
When rates are low, some (many?) investors may decide that the stock market looks better than lending (either directly or indirectly).
And then that can be a self-reinforcing cycle as well--more money going into stock markets == stocks perform better == stocks look like an even better investment.
That's the first (big) reason why PE is increasingly a weak Value (in a broad sense) signal. Another huge one is that companies who are free cashflow monsters that shunt cash towards some sort of internally compounding flywheel (ex: Amazon with AWS CapEx and warehouse buildouts) screen terribly on PE ratio even though the cash they are throwing out is being plugged into a compounding engine far greater than one can find in the broad market. That's a value in and of itself, and obviously demands a premium and is well understood today.
Personally, most of the more fundamental oriented stuff I read seems to focus more on figures like sales to EV when ranking "value" in a peer group, not PE.
That said, it can be useful. I grabbed some Abercrombie because it's at a PE of 8 and all of the free cashflow goes to buybacks. Mostly chose to discount growth and focus on near term value, so there's no tricky modeling there, and again the cash is just shunted into buybacks and there is no complicated debt situation either. In an "old school" value proposition like that, PE can work decently.
But I'm also in a Canadian gold miner which is a far stronger Value proposition, yet it will screen horribly on PE... Go figure.
The more I focus on growth the more I'm gambling on unknown future earnings. Sure I might miss out on NVDA, but whether I'm better off missing out on it depends entirely on how that company ends up doing over my time horizon.
Focusing on earnings is more predictable. That doesn't mean its better and it isn't to say that today's earnings directly predict tomorrows earnings, but it is much more tangible today and is less risky as an investment compared to focusing primarily on growth.
Edit: when considering growth rather than earnings, how do you distinguish the investment from a ponzi scheme? When I ignore earnings and focus on growth, or hype etc, I can't help but feel like I'm just trying to ride the wave and sell to the next schmuck before it fizzles out. With earnings, and ideally with dividends, I can at least point to a reason the company may be worth that value other than point at what others want to pay for it.
This isn't a small point... If you are involved in the equity market and miss out on names like this, you're almost certainly going to massively underperform. The vast majority of returns comes from a small slice of names. And again, if you want to properly do value investing, you have to precisely understand the growth component of value, as well as the weird debt and cap stack situations that usually come with value names (the market is efficient and they trade at discounts for a reason).
I trade full time, and my personal long term account has no stock picking. Value or otherwise. The medium term acct does, but not the decade+ timeframe one does not. So ultimately I don't really know what to tell you. Picking the next NVDA is practically impossible, yet missing out on such a name destroys your relative returns against an impossibly simply approach (index investing). Therefore don't try. Simple conclusion.
If you are strictly focused on a very limited view of "value" such as free cashflow, asset value, and PE (which ignores important aspects like debt maturities, industry cyclicality, quality of internal compounding, etc), then you're frankly directly competing with private equity. The names that stand out on these terms get bought by private equity, and the scraps of "value" are left on public markets. PE has 500,000,000,000 in dry powder currently, and much more efficient access to Capital markets than you do, with the ability to lever up those easy "value" tilted cash flows many times and immediately sell the debt on to pension funds and such. Trust me, if the value opportunity was truly there, they would take it. What you're seeing on markets is fairly priced on a risk adjusted basis.
Just use an up to date factor overlay on top of efficient equity beta exposure if you want diversified value that won't pigeon hole you in weird value traps. Frankly. Figuring that out is going to be much easier than cracking the value investing code and somehow beating the index. I remember an interview from one of the smarter firm owners that I've heard and he uses "short junk" (which generates extra cash to deploy efficiently) to slightly lever up his equity portfolio while giving him a broad based value tilt (it's long short portfolios all the way down). Over a long time frame something like that is going to crush any form of stock picking for the vast majority of participants.
Show your work here though. I'd point to P/E ratios over time [1]. That's a very common signal used when valuing a company and today those ratios are much higher than historic averages, nearing the ratios we saw before the dotcom bubble burst.
[1] https://www.currentmarketvaluation.com/models/price-earnings...
Valuing a company needs to take risk free interest rates and future growth into account, so it's difficult to compare across vastly different companies and time periods.
Alternatively, we could cherry pick a single company. How do you justify tesla's p/e relative to the other car companies? From where I'm standing it doesn't appear to make any sense.
Differences in any of those factors between companies or between time periods will determine why a company is valued the way it is. Interest rates are way higher now than a few years ago, and still lower than they were in the 90s. Some companies have cool business models that give them an edge, eg. Being able to generate lots of free cash, being able to invest cash at high returns, borrowing at lower rates, looking in customers, or making deals that will assure high revenue growth for the future.
I dont know much about TSLA but I don't think the market sees them as just a car company any more than Amazon was just a retailer.
Maybe those would be roughly analogous in a stable situation, but my read is that the big players in the markets today are chasing short term gains despite the signals. I'd point to the housing crisis as a recent example, the biggest banks and funds were massively over valuing real estate and real estate derivatives because they were blind to the actual risks and saw what appeared to be free money.
HFTs do not (for the most part) have any kind of “valuation” of stocks based on the fundamentals of companies. To a first approximation most HFT strategies are market making- they buy from people who want to sell at 9 and they sell to people who want to buy at 10 and every time they successfully do that round trip they make a dollar.
There’s more to it than that because of course there is. Markets are complicated and there are a lot of market participants trying to do lots of things. So sometimes they make use of their knowledge of market microstructure to try to detect and preempt big market moves (eg index rebalances, big portfolio unwind trades etc).
But don’t think HFTs have a perspective or care about the value of TSLA or whatever as a company. They basically want people to trade TSLA a lot so they make the vig every time they do. As some sibling has pointed out, they don’t want toxic/directional flow though, they want balanced flow because that means they keep making a clip whereas directional flow leaves them holding a lot of risk one way or the other because they are on the other side of those trades.
Quantitative traders do add information.
HFTs probably don't like trading with people that are adding information, that's why order flow from retail traders is valuable.
Literally called Toxic Flow in the industry.