> Softbank could pocket 2.7% of $70B ($1.89B a year) by just investing in an index.

This is totally untrue. Volatility in the short term -- one serious year down -- can sink you.

No sane actor treats S&P's average rate of return as the risk-free rate.

A serious economic downturn is likely to sink the fund entirely anyhow, so you might as well optimize for the case where that doesn't happen. This is one of the serious principle-agent problems in investing in general - equity holders and management aren't incentivized to preserve value for creditors, so they take more risk than is ideal for the overall capital structure.

Depending on how you define a "serious economic downturn". The one thing funds that invest in private companies has it that they can mark-to-model rather than mark-to-market. If it's not a prolonged downturn, private funds can much more easily ride out downturns than funds that invest in public securities and have to actually mark things in a realistic manner. Given the longer lockups of most PE funds, investors are also along for the ride.

I don't know more than the basics, my thought was if you pocket $2b/year and invest it, you have $13B+ in 10 years (plus your own $28b you've invested that I didn't include). So $40B to whether a storm every 10 years seems reasonable.

I think you are ignoring a lot of factors that come into play when you are managing assets on the scale of 70B$. It is easy to sink 1000$ in a minute into an index fund using Robinhood. It is not as easy when you are talking numbers on this scale.

99% in SPY, 1% in 80% SPY Put option should cover your downside.

I think you're going to find the 80% SPY Put option doesn't have liquidity for 700 million$ of options.

> Oversimplified napkin math for fun:

We are having fun!

I was under the impression fund money is typically locked for around 10 years or so.

Even the 10 year rate of return has variance. The risk-free rate of return for 10 years should pretty much be 10 year TIPS bonds.

Is there a strong reason as to why a 9.7% annualized return on the S&P is a correct assumption we should make about the future?

I feel like in the 21st century, a lot of folks have come to put a lot of faith in the "stocks in the long run" mantra.

But any stationary effect in the markets can be arbitraged away. If 9.7% long-term returns were guaranteed, wouldn't everyone just borrow 30-year loans on margin at 5% interest rate and invest it in S&P?

Edit: agree with sibling comment. If one loses 27% in the first year and makes 9.7% annually after that, they would be operating at net loss. So even if long-term gains were assured, volatility can still make investing on borrowed money unprofitable.

I've asked this question many times in many different formats and have never gotten a satisfactory response that's grounded in fundamentals.

I think it ultimately comes down to people trusting history to repeat itself.

How would an answer based on fundamentals even look like? Nobody can predict the future.

> How would an answer based on fundamentals even look like?

Start with demographics. Layer on productivity growth and you have a first-order approximation of national productivity. Figure gross margins and from that net margins, as well as average corporate tax rates–boom, you have the economy's profitability. Estimate a pay-out rate and you get a high-level return estimate. It's very approximate, but it's theoretically solid for any closed economic system. (Cf: natural resource inputs are not properly accounted for.)

Let's say you do this exercise and arrive at a discounted cash flow model that accurately predicts a 5-10% YoY appreciation in value for the S&P index.

How do you explain why this appreciation in value isn't arbitraged away, as suggested by GP? In his words, "if 9.7% long-term returns were guaranteed, wouldn't everyone just borrow 30-year loans on margin at 5% interest rate and invest it in S&P?" This would effectively drive up the current day price and eliminate any projected future growth in index price -- invalidating the hypothesis that the S&P price will continue growing in the future.

It's not arbitraged away because it's not guaranteed.

How would you account for central bank actions?

We can't predict the future with certainty, but we can certainly build models.

I'm basically saying I'd like a somewhat bottoms up approach to modelling the S&P index where the inputs can explain the x% YoY increase in price.

Investing in 1968, your S&P holdings would take until 1992 to regain their inflation adjusted starting value.

The S&P has not always generated positive returns relative to inflation.

Sometimes there are 30 year long dips.

That is demonstrably not true. 1968 was a positive year!

1968: +11%.

1969: -8.6%.

1970: +3.6%.

1971: +14.5%.

1972: +19.2%.

$1 of S&P500 in 1968 would be worth $1.23 in 1978, 10 years later.

There are large drops in some of the years, but they are usually followed by outsized rallies in the years afterwards.

Source: Moneychimp’s CAGR of the Stock Market calculator. http://www.moneychimp.com/features/market_cagr.htm

Inflation adjusted were the key words.

Even your link admits poor performance during periods of high inflation, and gives several other example periods.

But here was my original source:

http://www.multpl.com/inflation-adjusted-s-p-500

With your link, if you're just talking about value of the stocks and controlling for inflation, one dollar of stocks in 1969 grows to only $1.03 in December 1991.

9.7% (or maybe 0.8%) is the average over the last 90 years and I don't think there is any other data points you could reasonably use to say it won't continue like that over a long enough time frame.

Well, over the last 90 years the USA moved from important regional power to the economic and military hegemon of the world. That's quite the wave to be riding as a US stock market investor. It's also quite a hard act to repeat.

To maintain the historical growth rate, you have to believe one of the following two things, or some combination of them:

1. The US will continue to amass a larger share of the world's wealth, indefinitely.

2. The economy of the world at large will begin to also grow ~10% a year, a number far in excess of the historical average or any well-informed estimate.

Personally, I wouldn't make that bet.

Sure there are; valuations are considerably inflated now compared to the average (or starting point) of the last 90 years. Percentage earnings and dividend rate are both considerably lower. The long term expected return for the market is therefore also lower.

Taxes certainly complicate that, but most people do what you're describing. People with moderate and even high net worth (that is, in financial assets) still use mortgages because it's better to borrow at 5% instead of liquidating an asset that is returning 9.7%.

As others said, if 9.7% was guaranteed enough, who would lend Softbank at 7%? Banks aren't that stupid, they'd buy S&P (feedback loop)

Even 7% is an extremely good return rate.

Softbank does these things because they have access to a lot of easy money.

'very little risk' - I think standard deviation of return on S&P is something like 15%-20%...I wouldn't define a strategy with a buffer of 2.7% very little risk when the standard deviation is that high...

Pasted from another comment:

I don't know more than the basics, my thought was if you pocket $2b/year and invest it, you have $13B+ in 10 years (plus your own $28b you've invested that I didn't include). So $40B to whether a storm every 10 years seems reasonable.

Variance is not risk. https://www.ncbi.nlm.nih.gov/pubmed/18554271

No, but variance causes risk when you're borrowing to invest and need a certain level of return to continue servicing that debt.

If you had started such a scheme in 1999 or 2000 you would be bankrupt by now.

Not if you invested in an adapting basket of stocks (ie, an Index like the S&P 500) used dividends (2% to create a cash cushion) and had access to 0-2% short term debt (ie the Fed bank rate) during every downturn.

I was talking about taking $100 (at 01/01/99 or 01/01/00), investing them in the S&P500 (with dividends reinvested), and paying out $7 every Dec 31. You would run out of money eventually.

Of course if you add cheap additional leverage at certain moments and time the market successfully you can do better. But that’s not the original proposal.

If you think the S&P consistently returns ~9.7% per year every year, you are mistaken.

Pretty close actually. https://www.nerdwallet.com/blog/investing/average-stock-mark...

Edit: Anyone want to clue me in on what the downvotes are for here?

The Nikkei 225 is still down from 20 years ago. Its fate is a very plausible alternative history for the S&P.

Taking into account reinvested dividends, it's given an annualised return of 3.5% in the past 20 years. Source:

https://dqydj.com/nikkei-return-calculator-dividend-reinvest...

From that article:

"While 10% might be the average, the returns in any given year are far from average. In fact, between 1926 and 2014, returns were in that “average” band of 8% to 12% only six times."

So if you are investing on borrowed money with a guaranteed payout of 7% a year, that is not good for you.

The key is consistently.

The S&P 500 returns are wildly variant, and the 9~ % only comes from averaging a century of growth. Whole decades might see a loss.

Based on numbers here: https://seekingalpha.com/article/2789035-s-and-p-500-index-r...

1940-1950: 9.3%/y

1950-1960: 20.0%/y

1960-1970: 7.7%/y

1970-1980: 5.9%/y

1980-1990: 17.2%/y

1990-2000: 18.2%/y

2000-2010: -1.0%/y

So only decade with negative growth.

Add in the depression and you'll probably see a bunch more...

You need to consider inflation. In 1970, the inflation rate was 7-8%/y on average, so the "growth" these numbers imply did not exist in reality.

Good point. I think also debt’s effect is the perfect alignment for their “last” stage investing where risk and return are lower.

Aren't all of the 2.7% gains wiped out by inflation?