Bank of America: The S&P 500 Is Going to Hit 3,500 by the Year 2025

“Based on current valuations, a regression analysis suggests coumpounded annual returns of 8 percent over the next 10 years with a 90 percent confidence interval of 4-12 percent”

Regression analysis? Well, I guess it is better than just blanket stating 7 percent returns, like all these other people do.

http://www.bloomberg.com/news/articles/2015-11-24/bank-of-america-the-s-p-500-is-going-to-hit-3-500-by-the-year-2025

Cut-and-paste working again? Or did you switch browsers?

No, I actually typed the quote. I can paste things in the thread title, however.

Sounds legit where can i buy 10x leveraged S&P ETF?

You guys knew I was going to show up on this thread right? Oh this is fun… Mmmk. So let’s look at history. Since the 2007 peak of 1550, to 2015 of 2100, that’s only a 3.9% CAGR. But during this time the P/E has exploded to an ultra-stretched 23X. Why? Because the index went up faster than earnings, of course. Valuations are now at one of the highest points in history. Now these mathematically challenged individuals expect it to go to 3500 by 2025? Based on what earnings? With real GDP growth around 1.5% for the next 10 years, we are supposed to expect some wild earnings that will not only correct the current overvaluation BUT ALSO send the index to 3500, WITHOUT it being in a bubble? Let’s see the math on that please! Let me do it for you…a 9.9% CAGR on earnings, for ten years straight, during rising rate environment, with no recessions, such that you end at 3500 and 15X P/E. But S&P EPS growth for the last 10yrs was only 3.2%.So probability of this scenario, maybe like 5%? Buy yeah, let’s make that our base case. :wink: As Trump would say, “Savita Subramanian, you’re fired”.

stock markets can only go up though. its totally safe

Here’s a reasonable base case…1) EPS continues to grow at 3% rate, 2) interest rates rise and equity valuations come down to 18X, which is still significantly above the 100-yr average of 15.6X. Which means…

1.03^10 * $90.85 Q3 ttm EPS = $122.09 EPS x 18X = S&P500 2198 in year 2025

Ooops, can publish that. :wink:

^ For the next ten years, you’d take bonds over stocks?

PA you are saying over the next 10 years the S&P is only going up ~100pts? If only there was a RemindMe feature. Would like to see what really happens, but if that is correct I would assume you are out of US equities

Although to ohai’s article, it is pretty hilarious they really ran a regression to determine this. I have to imagine they know that is BS number

i’m not an outright bull but let’s add some realism to your estimate.

earnings are closer to 107 currently. you can’t just pick some random number that you think earnings is, you have to use a number that is actually comparable to the data of the last 100 years. as a result, the current P/E is ~19.5x, not the 23x you are assuming.

this P/E inherently tells us that the earnings yield is 5.1% currently with no earnings growth. let’s say earnings grow by the 3% you have assumed and that the S&P 500 pays dividends of 2% per year going forward so total earnings growth of 6.1%, which is logical because earnings that aren’t payed out are reinvested. $107 * 2.33 = $249.

corporate profits could probably fall about 30% from now until 2025, or they could stay the same, but let’s say they fall 30% to reflect a more normalized corporate profit % relative to GDP. so $249 * .7 = $174

with these assumptions and assuming the S&P trades at say 17x to reflect a still very low interest rate in 2025 compared to the last 100 years, the S&P would need to trade at 2958 in 2025.

please do 10 mins of research before spewing buckets of useless information all over the internet.

So 3500 is approximately a 75% growth over 10 years, which comes down to 5.8% capital growth average.

With 2% dividends paid, that’s about 7.8% total return average, which is fairly close to the long term average for stocks. Perhaps a bit high if you assume autarkic markets and structural changes in the economy. Not necessarily high if you assume that S&P companies are exposed to global growh.

Remember, though that this is a nominal return. So if inflation is 2%, then there only needs to be about 6% real earnings growth to support that. If inflation shoots up, as most people who claim that the US is falsifying data and bankrupt and debasing the currency believe, then those companies’ assets should track inflation and rise, even if they aren’t growing in real terms. It’s harder to say what will happen to the Present Value of Future Earnings portion of the value under inflation, but it is true that those who’s PVFE are helped or least hurt by inflation are more likely to work their way into the index, and those that don’t are likely to be dropped from the index.

None of this means that there can’t be a sharp pullback or crash in the interim, but 10 years is a while to smooth out those changes over time.

All in all, 3500 by 2025 is a pretty standard kind of projection, and they have a fairly large margin of error supplied to cover their asses too. What’s remarkable about this projection is how unremarkable it is.

who cares about EPS, annualized returns, CAGR, etc, etc

The variable that matters most here is the Fed. They could easily get the S&P past 3500 in no time with QE Infinity. And if the S&P is at 3500, just imagine where gold will be!

You are wrong. Q3 EPS (ttm) using as reporting the in actual financial statements earnings are what I said, $90.85. This number is from S&P’s Excel published this week and is the most up-to-date number we have (96% of companies have reported earnings).

You can’t just pick some random number that you think earnings are, you have to use a number that is actually comparable to the data of the last 100 years. That number is AS REPORTING TRAILING EARNINGS AS REPORTED BY S&P. And per S&P, “as Reported earnings represent the longest monitored earnings series available today”.

Dude you just got slaughtered. We should just stop now…

You are wrong. The P/E (ttm) using as reported earnings is 22.992515X as of market close Nov 25th 2015, which is not “an assumption”, it is the number reported by S&P, and there is a 100+ year history to which you can compare. There is no better number, this is THE authorative number, the proper calculation, and is not up for discussion.

Your math is flawed. Please recheck and find your error.

I suggest you follow your own advice, for example starting with: http://kr.spindices.com/indices/equity/sp-500

There is a real lack of deep analytical ability here. Front office I’m guessing. :wink:

The US 10-yr nominal GDP avg growth rate is 3.3%. Over this time the S&P500 has been able to do 3.2% EPS growth. Inflation has been 2.0% during this period. Returns have been 6.8%. The reason returns have been greater than EPS growth, is of course multiple expansion. Multiples expanded from the start of the period 18.5X to 22.9X. Multiple expansion is not sustainable. So when you say the 7% total return continues you are saying MORE multiple expansion, on an already expanded base, with no correction in sight. Because earnings growth sure as hell ain’t there.

S&P500 EPS Growth Averages (source: S&P), and Nominal GDP Growth (source: US Bureau of Econ Analysis):

25YR AVG 7.4% EPS, 4.5% GDP

20YR AVG 4.9% EPS, 4.4% GDP

10YR AVG 3.2% EPS, 3.3% GDP

When do the actual earnings happen? They don’t, because the US is a steady state, 1.5% GDP forward-growth economy, maybe 2-3% earnings growth max. It’s like NASDAQ and other bubbles, it takes years, sometimes decades, for earnings to catch up with bubble index values.

How again does 23X P/E hold during rising rates, falling GDP grow and falling EPS growth? Solve the equation. The mistake everyone is making is running backward looking numbers, and realistic forward looking numbers are nowhere to be seen. My forecast is for massive statistical manipulation, questionable monetary policies, etc, (in other words more of the same)…because apparent returns MUST happen or capital will exit the country.

Yup, I bought Chinese A-shares at 5-8X trailing P/E during the Aug crash, lots of room to run there over the next 20yrs. I’m short S&P500 at 2100, a dead end market.

Yes, when you edit out most of the analysis, it’s very easy can make others sound extra shallow. Congratulations. My ten year old stepdaughter figured this out too.

I admit that my analysis was pretty back of the envelope, but the point was to show that 3500 from here is not a huge stretch, since there is 10 years to do it, it’s a nominal figure that includes inflation, the s&p 500 is mostly global corporations (so US growth is not as relevant as you’d think), the s&p 500 by construction evolves to incorporate the winners in the profit gain (so will be biased to incorporate those with high EPS growth), and the error bars on the estimate are freakin enormous.

I think the issue with this analysis comes down to extrapolating a factor that cannot continue its trend for another 10 years:

@ pa. so let’s say we use the recorded TTM S&P earnings and don’t normalize earnings at all. that means you expect Energy companies to write off massive segments of their business EVERY YEAR! okay, so similarly at the market bottom in 2009, you would’ve concluded that we should use S&P earnings of 30 or so and that the banks were going to writedown a large part of their business EVERY YEAR and that the market was a terrible value as a result. where can i sign up to have you as my investment counsellor? i love your strategy of selling at market bottoms and buying at market tops. great idea.

you HAVE to normalize earnings after a single sector records massive writeoffs. you HAVE to. if you don’t, any further analysis is going to be wildly wrong.

the figures i chose are the figures that every investment bank in the world is using. i’m glad you’re using data that every uneducated moron trading out of his mother’s basement is using.

further, please don’t attempt to call bchad out on his numbers. if there is one guy on this forum that makes me nervous having to justify my work to, it’s bchad. you will be destroyed.

my math wrt to EPS growth is wrong. was clearly thinking of something else. but bchad is right. basically take your earnings yield, 5.1%, subtract dividends of 2.0% and add inflation, say 2%, and that, 5.1%, is your minimum EPS growth forecast. so with no change in valuations or corporate profits the S&P 500 would be at 3412 and you’d get 2% in dividends per year. with a decline in the P/E to 18x from 19.5x, you’re looking at 3150 plus 2% dividends per year and if we include the absolute worst case reversion, a 30% decline in corporate profits, and a P/E reversion, you’re looking at 2200 and 2% dividends per year.

your numbers are way off.

earnings growth is not as important as valuation increases.