QQQE

[Kingran]

I’m not confusing stock performance with valuation…Growth stocks dominate the top spots most of the 25 years covered, with the exception of several years after the tech bust in the early 2000s when Exxon had the lead.

Did I miss it? Where is the argument that mega cap’s are typically overvalued???

Kingran wrote, incorrectly: take any of your favorite valuation measurement and use it, you will be surprised to see the big cap’s are generally lot cheaper, be is PE, Price to sales, ROE or ROA, risk adjusted return.

I took an easy shortcut to show that was wrong, at least right now,

Again, respectfully, where is the data that shows megacaps are perennially overvalued?

Let me remind you the argument for the equal weighted is megacap’s are overvalued, hene you want to spread the bet on lower valued names.

So far I haven’t seen any evidence.

In this board, if I say 1+1=2, and you argued it is actually 4, you will get 100 rec, shame on me. But that still doesn’t make your argument is right. I am waiting to see the evidence.

I am not really good at running back-test, but there are others here good at it. So, if they can do a back-test that shows over 10, 15, 20 year period equal weighted index outperforms the cap weighted, I am happy to accept it. So far attack the person asking the question is what I have seen, If you cannot provide any evidence, I can understand. It is a long-list of assertions that have willing audience and damn the guy who is "challenging it’.

[rdutt]

So, if they can do a back-test that shows … equal weighted index outperforms the cap weighted, I am happy to accept it.

You can backtest over any time period and get the answer for free at portoliovisualizer.com.

Over the past 10 years, the regular S&P 500 (SPY) beat the equal weight version (RSP).

May 2012 - May 2022 (10 years) 
              Max
      CAGR    Drawdown
SPY   13.5%   -19%
RSP   12.9%   -27%

Not only did SPY beat RSP, the RSP investors suffered a much worse max drawdown (-27%) compared to the regular SPY investors (-19%). There’s no safety in equal weight.

What about QQQ versus the equal weight QQQE?

May 2012 - May 2022 (10 years) 
              Max
      CAGR    Drawdown
QQQ   17.5%   -22%
QQQE  14.5%   -20%

QQQ absolutely trounced QQQE over the past 10 years. $100K invested in QQQ in May 2012 would have grown to $508K today versus only $392K if invested in QQQE. The max drawdowns were again about the same, so there was no safety in QQQE either compared to QQQ.

“Just the facts, ma’am.”

[ZD1]

In this board, if I say 1+1=2, and you argued it is actually 4, you will get 100 rec, shame on me. - kingran

So all the ranting all the time is just about not being loved?

[Kingran]

So all the ranting all the time is just about not being loved?

Yeap, it is all about me, me, me… LOL

[mungofitch]

“Just the facts, ma’am.”

“Just the selective facts”?

Re the S&P versus the S&P equal weight, for example–
Yup, the S&P beat the equal weight in the last decade.
Not by a lot though–despite the roaring market for some supercaps, the cap weight won by only +0.40% in the ten years to June 17.

But more importantly–
Have you thought deeply about the reasons behind the short term result you report?
Which effects are the cycle, which are structural?
Which can be generalized and extrapolated, and which should be seen as ripe for mean reversion?

For the conventional float-adjusted cap weighted S&P 500 versus the equal weight version,
the odds of the conventional S&P beating the S&P in a ten year period are 21% based on history.
(using daily dividend- and inflation-adjusted daily data from 1930).
So, it definitely happens. Just not often. So, it’s not what one should have as a central expectation.
The S&P 500 didn’t win over the equal weight in any rolling 10 year interval with daily end dates from mid 2001 to mid 2019.

In the daily-start rolling ten year stretches since 1930, the equal weight won an average of by 2.13%/year.
In ten year stretches following a decade that the S&P did better, the equal weight won by an average of 2.70%/year: some mean reversion.
So, knowing nothing, one ought to expect the equal weight to do better.
Knowing that the cap weight has done better in recent years, you should expect the equal weight to win by an unusually large amount.

What about risk?
The max drawdown figures are not useful proxies for risk, of course.
The only purpose of investing in an index is to diversify away from company specific risk and having to know the valuation and prospects of specific firms in detail.
So the cap weight indexes are far riskier, since they have around 6-11 times the company specific risk.
Risk isn’t short term price squiggles, risk is losing money. Or not making the money you needed.
Using a more sensible metric of risk, rolling ten-year real downside deviation with MAR=8%,
the risk of owning the equal weight is less than 59.84% of the risk of owning the S&P 500 and its cap-weight predecessors.
(This is a measure of how often returns fell below inflation + 8%/year in any ten year stretch, with a squared penalty on the size of the shortfall below that figure)

But, speaking of drawdowns–this recent result is a wonderful demonstration of the randomness of the S&P 500 because of its high concentration risk.
Think about why the max drawdown was lower with SPY in the specific stretch you looked at. Is it meaningful?
A cap weight index has a huge concentration in a few names.
In a panic, almost everything sells off, but not everything.
Sometimes the gigacaps are unusually bad performers during that stretch, sometimes unusually good.
It’s pretty random–they’re only a few companies, after all.
We recently had a pandemic, and the ones that happened to be at the top at the time were seen as winners during lockdowns and did very well.
The best way to think of the returns of the S&P 500 are like this:
The returns of the S&P 500 equal weight (what the typical firm is managing),
minus a small constant,
plus or minus a big random number depending on the luck of who’s biggest and overweighted right now.

A possible moral for the story:
Never mistake the cycle for the trend.

Jim

[bigshan]

In the daily-start rolling ten year stretches since 1930, the equal weight won an average of by 2.13%/year.
In ten year stretches following a decade that the S&P did better, the equal weight won by an average of 2.70%/year: some mean reversion.

I have a feeling this can’t be explained by statistics (expected return, not the probability of very bad return) unless smaller-cap inherently outperform bigger-cap.

[mungofitch]

I have a feeling this can’t be explained by statistics (expected return, not the probability of
very bad return) unless smaller-cap inherently outperform bigger-cap.

Yes and no.

It’s not so much that “smaller firms do better on average”. They don’t, really. Not by that much, anyway.
A better characterization is that “the stock prices of the few truly giant firms tend to do particularly badly”.

So how does equal weight outperform?

First, there is a small effect from the pure rebalancing process itself.
Consider a simple case: if all stock prices move by a random amount each quarter but were otherwise flat long run,
a cap weight index would be forever net flat and an equal weight one would have a small long run profit.
Each quarter the equal weight index would sell a bit of something that (randomly) rose in price the prior quarter,
and buy a bit more of something else that (randomly) fell in price. The next quarter those returns would (on average) reverse.
This price rebound effect isn’t that large, though…it can be measured.

Rather, the biggest contribution simply comes from avoiding an over-concentration in the very few very largest firms by current market value.
They are, statistically and for fairly clear reasons, more likely than the average firm to be overvalued at any given time.
(Not always leading to underperformance, just on average over time)
The problem is multiplicative: these are usually the worst performers, and a cap weight index has a very large allocation to them.
Imagine reading the prospectus of a fund that described its strategy that way–would you invest?
An equally weighted portfolio of the five largest US-listed firms by market cap 1997-2016 returned 4.05%/year.
The S&P 500 returned 7.68%/year.
The S&P equal weight returned 9.47%/year.

That’s why there is nothing particularly magic about the equal weight strategy.
A huge part of the benefit can be accomplished by using any reasonably diversified weighting scheme based on a metric other than market cap.
So, it’s not about moving to equal weight, and it’s not about smaller firms doing better than larger firm. It’s about “anything but cap weight”.
Have a look at PRF using the RAFI methodology. It weights positions based on the size of the business, rather than the market cap of its stock.
They are similar in practice–PRF’s current largest posistions are Apple, Microsoft, JP Morgan, Berkshire and Exxon.
But there is no overweighting of a position merely because its price goes up.
(they measure company size based on proxies like revenues, cash flow, book value, and total dividends paid)
Long run performance of their index has been good, though not exciting, despite the tendency of long
run underperformance of asset heavy businesses resulting from including book value in that list.

I note that the net five year returns for PRF (concentrated in big firms but NOT based on market cap)
and for RSP (equal weight, so smaller average business size) are virtually identical, despite shorter term variation.
Of the hundreds of different schemes that have been used, the outlier for long run performance isn’t equal weight, but cap weight.

Remember that cap weight indexes were not designed as investment vehicles.
They were designed as information tools for watching the progress of markets.
Only later did someone suggest investing in them as such.
The attraction for the fund company is obvious: almost no trading.

Jim

[inparadise]

Wow, Jim. You are patient. So very much appreciated.

IP

[bigshan]

Jim, thanks for the detail explanation, it makes sense.

[bigshan]

I would like to add that besides better return long term, the equal weight also has lower risk of very bad return. Suppose Tesla went to zero and the cap-weighted index had 5% of the index in Tesla, the the index would lose 5%, while the equal weight would only lose 1%

[mungofitch]

Wow, Jim. You are patient. So very much appreciated.

They’re very complicated issues.
I have the tailwind of some very good data sets.

Note, nothing in what I’ve said suggests that the current super-large firms are going to do badly as business, nor even that their stocks will do badly from here.
I’m not even arguing that QQQE will beat QQQ or that RSP will beat SPY in the next few years.
Maybe yes, maybe no.
The outcome of the race depends on the fortunes of a very small number of very big firms.
But that’s rather the point–
The cap-weight returns should be thought of as the somewhat predictable equal weight returns, plus or minus an unpredictable random number.

My stance is this:

If you know this year’s huge giants well and can value them and make a sensible estimate of their likely returns, invest in them, not a cap weight index.
You’re not the target audience for index investing.

If you ARE the intended audience for index investing and don’t know how to value them or their individual prospects,
don’t make a hidden huge concentrated bet on them with a cap weight index.
On average the giants of the day are substantial underperformers, so it’s not what you want to bet heavily on.
And even without that knowledge, you’re not a stock expert, so you don’t want to be big on anything.
Take a choice that has the vastly lower risk and usually better returns.

There are many such choices, not just a fund that gives you everything equal weighted.
Personally, I like the “monkey with a dartboard” approach.
Pick 30-50-100 stocks at random that are included in a major index, buy equal dollar amounts of each.
Hold each position for a year or two, selling each at a predetermined time, and replace it with something else.
Your chances of beating the S&P are very very good.
And you can eliminate from consideration a few firms you might find odious, so it is more ethically defensible than index investing.
If your dartboard happened to be just a little bit biased towards the 10% of companies with the highest ROE, it probably wouldn’t hurt : )

Jim

[mungofitch]

I would like to add that besides better return long term, the equal weight also has lower risk of
very bad return. Suppose Tesla went to zero and the cap-weighted index had 5% of the index in Tesla,
the the index would lose 5%, while the equal weight would only lose 1%

This reasoning also works in mild cases.

For example, I think Microsoft is probably a bit overvalued at present. Maybe I’m wrong about that.
But if it’s true, then maybe they will have no return for the next 2-4 years, even if the business continues to do well.
Microsoft is over 10% of QQQ and 5.6% of SPY.
But of course only 1% of QQQE and 0.2% of RSP.
Would you rather have a flat return from 10% of your portfolio because of the idiosyncrasies of a single stock, or 1%?

On any given day, some stocks are temporarily overvalued, to widely varying degrees.
On that day, every cap weight index is overweight every one of those stocks, and underweight all the undervalued ones.
The overweight of each is in proportion to the degree they are overvalued that day, times the size of the firm.

So, the main drag is big firms that have unreasonably high prices.
Tesla was quite richly valued when it was added to the S&P in December 2020 replacing AIV.
The S&P 500 index was an amazing 0.41% lower six months later than it would have been without that single replacement.
The drag from that event in the equal weight S&P index was under 0.05%.

For me the moral is this:
Never take big positions in things you don’t have a good reason to want a big position in.
That reason can only be a justified expectation of some mix of above average prospective returns from current prices and below average likelihood of permanent capital loss.

Jim

[earslookin]

Pick 30-50-100 stocks at random that are included in a major index, buy equal dollar amounts of each.
Hold each position for a year or two, selling each at a predetermined time, and replace it with something else.
Your chances of beating the S&P are very very good.

Equal weight portfolios of 50-100 stocks picked at random from an index will result in a normal
distribution – about half will do better than the index and about half do worse. Results will
also depend on how the major index itself does vs the S&P.

Ears

[PhoolishPhilip]

This has been a fascinating and informative conversation.

If you ARE the intended audience for index investing and don’t know how to value them or their individual prospects, don’t make a hidden huge concentrated bet on them with a cap weight index.

This point resonated with me particularly well. I spend some time re-balancing my 403-b last fall because I was concerned about the heavy concentration in a few big name stocks in my funds. For example, the top five holdings in VINIX are Apple, Microsoft, Google, Amazon, and Tesla for a total of 20.6% of the fund. It was even higher last fall before prices started falling. I was not comfortable with this level of concentration, but my company does not offer a ton of equal weight options. I ended up having to shift to mid cap and international funds to balance out the concentration risk.

PP

[Kingran]

despite the roaring market for some supercaps.

Hmmm I thought everything was overvalued, now the narrative is only supercaps were overvalued??? Selective reading?

The max drawdown figures are not useful proxies for risk, of course. The only purpose of investing in an index is to diversify away from company specific risk and having to know the valuation and prospects of specific firms in deta

Why not? Of course drawdowns are not measure of risk, because it is quantifiable and used by everyone in industry.

The only purpose of investing in an index is to diversify away from company specific risk and having to know the valuation and prospects of specific firms in detail.
So the cap weight indexes are far riskier, since they have around 6-11 times the company specific risk.

Hmm pretty faulty logic. No one needs to know the valuation or risk profile, even if they are cap-weighted. Arguing against straw-man.

But, speaking of drawdowns–this recent result is a wonderful demonstration of the randomness of the S&P 500 because of its high concentration risk.
Think about why the max drawdown was lower with SPY in the specific stretch you looked at. Is it meaningful?

Again, why not? May be the market assigns higher valuation to the biggest firms and ones with less risky. But is this argument even true?? NO. Here are the big caps’ that suffered far higher drawdown than the Index itself!!!

AMZN	GOOGL	TSLA	META	SPY
189	3030	1243	384	480
110	2248	680	164	381
				
42%	26%	45%	57%	21%

Selective reading, faulty logic. It is still not clear that data is convincing, rather, data seems to be selected based on one’s hunch.

If the equal weight provides outsized return or sustained advantage, why is that product not popular? After all the fees are higher for that.

The best I could see is, the differences are on a given year is low, but over time could add up. Will I choose a strategy, without clearly understanding why it is superior, and pay higher fees, and suffer lesser liquidity? I don’t know.

But there are many who likes passionate, often very wrong, arguments. Good luck.

[Kingran]

Suppose Tesla went to zero and the cap-weighted index had 5% of the index in Tesla

Before it goes to zero, the weightage would automatically go down, do you realize that?

[said2]

Jim: The only purpose of investing in an index is to diversify away from company specific risk and having to know the valuation and prospects of specific firms in detail. So the cap weight indexes are far riskier, since they have around 6-11 times the company specific risk.

King: Hmm pretty faulty logic. No one needs to know the valuation or risk profile, even if they are cap-weighted. Arguing against straw-man.

Faulty logic?

A) Investing in single companies => Company specific risk. Imagine having 100% of your portfolio in one company. Handling that requires knowing the valuation and risk profile of that specific company - which logically is what correctly Jim said when he inverted that and said the purpose of investing in an index is the opposite.

B) Cap weight indexes by definition are dominated by the largest companies => That’s countering and contrary to diversifying away from single companies, sabotaging that purpose of an index, as you again need to know the valuation and risk profile of those few companies dominating the index to assess the index as a whole.

But there are many who likes passionate, often very wrong, arguments.
Seems I am one of them, one of those poor souls who often rec Jim’s posts with their faulty logic.

[bigshan]

Before it goes to zero, the weightage would automatically go down, do you realize that?

It’s all depending when do you buy in. The difference between buying when Tesla is at the top and buying when it’s at the bottom is significant. I think statistics over long term would cover those up and down events. There is another subject: since QQQE requires frequent balancing, selling more of the winners and buying more of the losers, the capital gains would pass to investors at year end and requiring paying more tax, right?

[mungofitch]

Pick 30-50-100 stocks at random that are included in a major index, buy equal dollar amounts of each.
Hold each position for a year or two, selling each at a predetermined time, and replace it with something else.
Your chances of beating the S&P are very very good.
…
Equal weight portfolios of 50-100 stocks picked at random from an index will result in a normal
distribution – about half will do better than the index and about half do worse. Results will
also depend on how the major index itself does vs the S&P.

Yes, but one clarification is needed:
Equal weight portfolios of 50-100 stocks picked at random from an index will result in a normal
distribution – about half will do better than the equal weight index…

Which is why it’s almost guaranteed to outperform the cap weight index.

I’ve read the papers, and redone the tests myself. The returns are good.
As you note, they’re tightly clustered just above and below what an equal weight index will give.
It’s a fair bit of typing once every year or two, and you have to buy a dartboard.
But the fund management fee is zero, and you have the control to skip firms that make (say) nerve gas if you like.

As yoy note, it does also depend what index or list of eligible stocks you start with.
If it’s the S&P 500 list your returns will cluster very tightly around that of RSP.
FWIW, an equally weighted portfolio of Value Line stocks, reset to equal weight each 3 months,
beat the S&P 500 by 4.26%/year in the 22 years to 2022-Q1.
That’s one main reason I use their list of 1700 as my usual hunting ground.
The S&P 1500 set would presumably give similar results.
The Value Line set seems to include a small number of hand picked very small firms they think have interesting prospects.
Maybe 25-40 of them, depending on how you define “very small”.

Jim

[mungofitch]

the capital gains would pass to investors at year end and requiring paying more tax, right?

For some people I presume so.
Though the difference is likely to be the time value of the tax, not the tax itself.
Presumably both options are profitable over time and will eventually have a tax bill due.

Jim