Hacker News Reader

Comment by kragen | Hacker News Reader

I think you're being a little unfair with the S&P comparison: you should be comparing the returns on your actual investment with the expected returns on the S&P, not its actual returns. The stock market got a 5.4% return over that time, yes, but that's above average; we should expect its long-run returns to be close to the rate of economic growth, or more like 3%. And it could easily have been zero or even negative over a short time like 13 years.

By comparison, the risk that your solar panels will stop producing electricity, or that someone will invent Mr. Fusion and drop the price of electricity through the floor in the next decade, is quite low.

View on HN

True, but its too easy to get lost in financial modeling. There were two high order bits in my decision, first was that the money was spent either way; Either I would be paying PG&E month to month, or I'd be paying off the system. And there was a good shot that the panels would generate as much or more energy for that investment. The second was that I felt it was the right thing to do, not a lot of people were in a position at the time to pay those sorts of costs and I really believe that finding better ways to power the world was a noble cause, by investing in a solar system, even if it was a 'bad' investment, it was jumpstarting the market for solar. That helped with the chicken/egg problem of people investing in new solar tech to meet a market that was (by the action of early adopters) clearly emerging.

So every time I find myself having this discussion in a group, there are folks who want to argue the financial model to "show" something (generally negative) about solar power. Either it's not a good investment, or it's just coal power used to make the stuff that now is generating not even as much power was was used to manufacture it, or it's causing more pollution by forcing utilities to turn plants on and off, etc etc.

I was fortunate that I could finance it without risking things like my kid's college fund, I live in an area where there is a lot of sun and mild weather so my house's "demand" is quite manageable, and as an engineer I get to live my belief that we can engineer solutions to the worlds problems like energy interdependence and climate change.

I am grateful that today, the systems costs have gotten so low that it's possible for more people to make that choice.

Just to be clear, I think it was a decent investment then (better than the stock market) and an excellent one now. I'm not one of the folks trying to show something negative about solar power.

I find your belief inspiring.

Dont forget your stock gains would have been taxed

May I ask where you got your 3% because that number is very low compared to any number I have ever read? This Wikipedia chart shows annualized returns to be 10.47%. That is >3x your 3%. https://en.wikipedia.org/wiki/S%26P_500_Index#Annual_returns

3% is an estimate of the rate of economic growth: http://www.tradingeconomics.com/united-states/gdp-growth As I said above, we should expect the market's long-run returns to be close to the rate of economic growth.

It's true that stocks in the US in the 20th century dramatically outperformed that 3%. This is for three reasons: first, up to about 1975, the world's economic growth rate was higher than 3%, due to the Second Industrial Revolution; second, up to about 1975, the US's economic growth rate was even higher than the world's, because it was taking over the markets previously supplied by the bombed-out economies of Europe (especially the UK) and Japan. Third, since about 1970, the share of economic output that went to owners of capital (rather than labor) increased dramatically.

Now, the third of those things probably can't happen again, because capital's share of output is already super high. (It's not that it can't get higher, but it can't go past 100%.) The second could happen again, but if it does, the US will be in the position of the UK, Japan, or Germany — it's now the Single Superpower (not even a mere Great Power) who could lose market share, probably to China or to a non-nation-state power. Investors in Japan's and Germany's stock markets in 1925 or 1935 didn't do so well.

The first, well, that's anyone's guess. It's a total wildcard.

It's certainly reasonable to posit a Third Industrial Revolution, made out of some of free software, nanobots, biotech, abundant solar energy, AI, and automated fabrication. I sure hope we get one. But if we do, it's not at all obvious who the returns will flow to. (Everyone, I hope, not just shareholders.) Accumulating capital goods, as we've been doing for three million years, is less important when your capital stock can double every 24 hours through self-replication. Breweries do not account for the quantity of yeast they have on hand as a durable capital asset.

And it seems equally plausible that we'll instead experience a new Bronze Age Collapse or Decline and Fall, as inexpensive DIY drones, very affordable precision projectiles, anonymous markets in assassination, and ubiquitous retail surveillance provide a decisive advantage for attackers over defenders in the physical world, just as their software counterparts have on the internet.

I understand the rate of GDP growth, the person you replied to specifically stated that they would have made more money on the SPX. Granted the S&P 500 is not guaranteed to go up, but historically it does.

"...first, up to about 1975, the world's economic growth rate was higher than 3%, due to the Second Industrial Revolution; second, up to about 1975, the US's economic growth rate was higher than the world's, because it was taking over the markets previously supplied by the bombed-out economies of Europe (especially the UK) and Japan."

Do you read the posts you reply to? Because the chart I linked starts in 1970. And you are discussing GDP when the post you were replying to specifically was talking about investing in the market.

There's not much reason (and even less empirical evidence[1]) to believe that U.S. stocks should follow U.S. economic growth rates.

[1]: http://www.economist.com/blogs/buttonwood/2014/02/growth-and...

To summarize the article, the long-run correlation between stock prices and economic growth rates is positive at 0.51, and in the short run the correlation is actually negative (no number given), because investors bid up stock prices in fast-growing economies, because they expect the growth to continue.

If the growth did continue, then they would make money, but often it doesn't, so they overpaid, so they get below-average (and often below-break-even) returns.

Also, per-capita GDP growth is irrelevant; what correlates positively with equity returns is total GDP growth, including the change in population.

So, on the contrary, this article provides a great deal of reason and even empirical evidence to believe that returns on US stocks should follow US economic growth rates.

I'm not an expert here, but doesn't it make sense for stocks to return higher than growth because of earnings?

For example imagine I own stocks making up 1% of in a company w/ earnings of $X, and market cap of 10 * $X. If the companies growth over a year is 3%, at the beginning of the year my stocks should be worth 0.01 * $X and at the end of the year 1.03 * 0.01 * $X due to the growth, BUT shouldn't i also have received 0.01 * $X (my share of earnings) in dividends that I am free to reinvest, giving me a 4% return overall?

If the company's value increases by 4% over the year, and it pays out one of those four percentage points as a dividend and reinvests the other three, you get the scenario you're talking about. (And it doesn't matter how much of that 4% is revenue minus cost of sales and how much is, say, increase in value of its holdings in another company or real estate.) But if the economy grew by 3% and the company's value increased by 4%, your company is doing better than the economy as a whole, and it's because someone else is doing worse. In the US over the last 40 years, a very significant "someone else" here has been the employees — as they have lost bargaining power, they have gotten worse bargains.

It seems like maybe you're saying that the revenues, and therefore the profits, don't count here because they're not part of economic growth, but just part of the ongoing economic activity. But what determines the P/E ratio, which is to say the return on capital (as a reciprocal), across the market? What keeps investors from bidding the market cap of the company up from ten times earnings to a hundred, a thousand, or ten thousand times earnings? It's the availability of other investments that they expect to grow in value at a higher (risk-adjusted) rate. If you can get a 16% annual return on your investment by buying solar panels instead of stocks, then the guy who does that will have 16% more money to invest in more solar panels every year, until either he bids the price of solar panels up (due to limited manufacturing capacity) or he bids the price of electricity down (due to limited transmission grids or electrical demand). The first of those is already happening; the second one should start happening in about 2024, earlier in some areas.

Now, you could argue that there's a difference between this solar panel maniac guy spending 16% of his capital base in solar panels every year, accumulating more and more solar panels and selling the electricity from them to buy more, and GDP growth, because the solar panel maniac is accumulating a stock, while what the GDP measures is a flow.

But note that by the hypothesis that the maniac is investing to get some relatively inflexible percentage return on his investment, he receives a flow of earnings that is proportional to that investment. And that flow is part of the GDP.

GDP growth has no strict correlation to stock market returns. Why not? Because corporate profits (eg the S&P 500) can grow far faster than GDP, as they did from 2002-2014.

You should not expect stock market returns to mirror GDP growth. The S&P 500 is not strictly representative of US economic performance, which is precisely why the huge corporations did so well with the cheap dollar and significant global economic expansion while domestic locked companies didn't fair nearly so well.

Your point that the S&P 500 represents large US companies, rather than all US companies, is well taken. Still, the process of centralization you describe (where most gains go to past winners) can't continue indefinitely; the S&P 500 companies (at least the ones in the US) can't account for more than 100% of US production.

Returns to the S&P should beat GDP growth even in the long run, steady-state.

GDP growth flows to both debt and equity. The debt market is actually much larger than the equity market and since its returns are lower than equities, and they both share GDP as a source of returns in the long run, equities should beat GDP growth in the long run. This argument isn't iron clad, but hopefully you can see the general picture.

The Gordon growth formula is taught in intro to finance classes to estimate returns to equity. The theory is that stock market returns equal the dividend yield plus growth in stock prices. Stock prices alone are in the long run expected to grow at the same rate as GDP, but the actual return should be higher by the rate of the dividend yield.

These are the most basic arguments against what you're saying, but I would also caution against assuming that any particular macroeconomic trend will end, especially that it will end in a timeframe relevant to decisions related to solar panel installations. Sometimes they just keep going. There is no mathematical reason that equity prices can't go to infinity, as required rates of return can decrease to zero. If the world becomes more predictable and stable, equities can keep becoming more and more valuable with no damage to finance theory.

Additionally, you'd have to subtract the monthly power bill from those returns. I.e. you can't say "I would have doubled my $20k return by investing it for 10 years", because you can't defer your monthly power bill for 10 years and pay it in a lump sum after you doubled your money in the S&P. Paying $100+/month for power would substantially impede the interest compounding effect.

No, that would amount to double-counting the returns. Yes, you need to time-discount future returns, and you may need to risk-adjust, but that's already taken into account.

The rate of economic growth has little relation to the stock market yield. For one you have forgotten about the dividends, currently around 2%. Now lets assume the company grows by 3%, so your shares will now be worth 3% more. But the total return will actually be 5%.

However here I assumed that there was no drop in optimism, or pessimism over the one year period. For if you were to bought in when expectations were high, your actual return would be lower.

Sometime the relation between growth and return is completely paradoxical. You would expect that in the past 100 years a portofolio invested in UK stocks to be dominated by one of US stocks. After all the first country went from being The Global Superpower to dubious second rate global player, meanwhile the US had done the same in reverse. Yet an investor in UK stocks would have been marginally better of.

Likewise you would expect China with its spectacular growth, to have brought impressive returns compared to the US. Yet that had not happened.

You aren't the only one to make those mistakes; I've explained where you went wrong in https://news.ycombinator.com/item?id=11623817 and https://news.ycombinator.com/item?id=11623377.

Actually the markets are "supposed" to double every 10 years, which is said that on average over 10 years, the markets will grow 10% per year. Since the recession 8 years ago that has been much lower, but in general, that is the rule most industry people will give you. I work as a tech guy in finance and have for the past 8.5ish years.

Anyone who tells you to expect a 10% annual return from the S&P500 is lying to you. And probably trying to sell you something.

Really, I do not understand your reasoning since the S&P 500 has an annualized return of >10% since 1970.

https://en.wikipedia.org/wiki/S%26P_500_Index#Annual_returns

Just because the S&P has seen such returns over the last 40 years, doesn't mean you'll see those returns over the next 40 years. The US economy is aging out, has stagnant wages, etc.

Also we saw a huge drop in interest rates that won't happen again over the next 40 years.

Going to disagree with you here. I think it'll be impossible to raise interest rates much further without causing economic damage. See negative interest rate policies in Japan and Europe, and close to zero rates still in the US.

I 100% agree with you about the difficulty of raising rates. But I don't think they'll continue to drop like they have over the past 30 years.

I have a hard time imagining a world of -5% to -10% rates!

> I have a hard time imagining a world of -5% to -10% rates!

A year ago I would've agreed with you, but with the ECB at -.04 and Japan at -0.02, I believe it more likely than previously.

Heh, fair enough. It's certainly very surprising to see rates below zero. So, indeed, I am much less certain about the future!

Incorrect, please re-read my statement. An average 10% annual return over 10 years. As the other reply to this notes, history shows in general this rule has held since 1970ish.

The U.S. markets reflect the U.S. ascending to become the dominant world power over the course of a century. Naturally, the markets grew during this period of tremendous economic growth. The investment industry likes to extrapolate from this, and declare it a rule of the markets as if it's a law of physics.

If growth slows or stalls, eventually the markets will catch up. A market that grows independently of the underlying economy doesn't make sense in the long term.

For something to double in 10 years it only has to grow by 6.9% per year. This is from a somewhat unusual application of The Rule of 69. Latter is way of determining doubling time of something that grows by x%: just divide 69 by x and that will be the doubling time in periods.

So if the DB grows by 10% weekly, then it will double in 6.9 weeks. If the interest rate is 3% then the debt will double in years, etc.

The Rule of 69 is good for a approximation, that works well with low single digit increases.

Reply on news.ycombinator.com