馃攳
Thematic ETFs (are Terrible Investments) - YouTube
Channel: Ben Felix
[0]
- Some investment themes are
more exciting than others.
[3]
Recently, themes like
crypto and blockchain,
[5]
marijuana psychedelics, cloud computing,
[7]
electric vehicles, clean energy,
[9]
robotics and artificial intelligence
[11]
have caught the imaginations of investors.
[14]
But this isn't the first time
[15]
that investors have clamored
over exciting companies.
[18]
Electronics in the '60s,
[19]
the NIFTY 50 in the '70s,
[20]
biotech in the '80s
[22]
and the internet stocks in
the '90s are good examples.
[25]
Similar examples go
back hundreds of years.
[27]
The idea behind these
eye-catching investment themes
[30]
is that investors can participate
[32]
in potentially disruptive trends
[34]
or innovations to earn excess returns.
[37]
As S&P describes their thematic indexes,
[40]
they offer access to a series
[41]
of technologically enabled,
often disruptive industries,
[45]
generally referred to in aggregate
[47]
as the Fourth Industrial Revolution.
[50]
I'm Ben Felix, portfolio
manager at PWL Capital.
[53]
I'm going to tell you
why thematic investing
[55]
is like lighting your money on fire.
[58]
If you're excited about an investment,
[60]
it's probably not a good investment.
[62]
The collective excitement of investors
[64]
has a tendency to drive up stock prices.
[67]
Few things are as exciting
[68]
as new technologies or industries.
[70]
I used to read about hyped
up investment crazes in books
[73]
but in my relatively short tenure
[75]
as a portfolio manager,
[76]
I have witnessed the phenomena first hand.
[78]
Do you remember marijuana stocks?
[80]
Everyone was going to be
rich until they weren't.
[83]
One of the big challenges
with exciting industries
[85]
is that they're exciting.
[87]
When people see a big potential market,
[89]
lots of businesses will form
[90]
to meet the potential demand.
[92]
The excitement about a big market stems
[94]
from profit opportunities for businesses
[96]
but however large the expected profits
[98]
from a new investment theme may be,
[100]
what matters to investors
[102]
is growth in earnings per share.
[104]
If lots of shares are issued by new
[106]
and existing companies pursuing
profit opportunities related
[109]
to an investment theme,
[111]
it is possible that the investment theme
[113]
will end up underperforming in the market.
[115]
This idea was formalized
[116]
in a 2003 paper by William Bernstein
[118]
and Rob Arnott titled "Earning Growth:
[120]
The Two Percent Dilution."
[122]
Even if there is a huge
pile of potential earnings
[125]
up for grabs based on a new technology,
[127]
the growth in earnings per share
[129]
of companies pursuing it
may not be attractive.
[132]
Most entrepreneurs and investors
[133]
don't think like this though.
[135]
Entrepreneurs starting
businesses in big markets expect
[138]
to be successful.
[139]
And investors selecting
companies to invest in
[142]
in big markets expect to pick winners.
[144]
The combination can drive up the prices
[146]
of companies pursuing
a hot investment theme.
[149]
Eventually, as investors learn more
[151]
about the market and the
level of competition,
[153]
prices tend to fall
[154]
and they often fall dramatically.
[157]
This dynamic of big markets
creating high prices
[159]
for companies despite limited opportunity
[161]
for investment returns,
[163]
followed by a drop in
prices when reality sets in
[166]
was described in a 2020 article
[168]
by Brad Cornell and Awath Damodaran
[170]
titled "The Big Market Delusion:
[172]
Valuation and Investment Implications."
[174]
They explained that the extent
[176]
of overpricing in a market
[177]
will depend on the degree
of overconfidence exhibited
[180]
by entrepreneurs and investors
[181]
where more overconfidence
leads to higher prices.
[184]
The potential size of the market
[186]
where a larger potential
market leads to higher prices.
[189]
The degree of uncertainty
about the future profitability
[191]
of the market where
greater uncertainty leads
[193]
to higher prices due to an amplification
[195]
of the overconfidence effect
[197]
and the perception of
winner-take-all markets,
[200]
again amplifying the effect
of overconfidence on prices.
[204]
Cornell and Damodaran
offer empirical examples
[206]
of e-commerce in the 1990s,
[208]
the online advertising
industry from 2015 to 2020
[212]
and cannabis stocks in 2018.
[214]
These were all big potential markets
[216]
with intense competition,
[217]
extremely high stock prices
[219]
and an eventual sobering drop in prices.
[223]
The effect of overconfidence
on asset prices
[225]
is thoroughly documented in a 2003 paper
[227]
by Jose Scheinkman and Wei Xiong titled,
[230]
"Overconfidence and Speculative Bubbles."
[233]
The authors explain that due to the cost
[234]
of short selling, overconfident investors
[237]
will tend to have more
of an impact on prices
[239]
than pessimistic ones.
[240]
And overconfident
investors will be willing
[242]
to pay more than their own assessment
[244]
of a stock's fundamental
value to acquire it
[247]
because it gives them the option
[248]
to sell to an even more
overconfident investor later.
[251]
The result is bubble-like
behavior in asset prices,
[254]
which doesn't tend to
end well for investors.
[257]
This is a real problem,
[258]
especially for naive investors
with limited knowledge
[260]
about asset pricing.
[262]
In a related 2013 lecture,
[264]
Scheinkman explains, "Experts that wish
[266]
to signal their familiarity
with new technologies
[268]
have a tendency to exaggerate their value
[271]
and in this way generate over optimism
[273]
among naive investors."
[274]
That over optimism among naive investors
[277]
is something that
financial product providers
[279]
are more than happy to cash in on.
[281]
A 2021 paper by Itzhak
Ben-David and three co-authors,
[284]
titled "Competition for
Attention in the ETF Space"
[287]
finds that specialized ETFs,
[289]
which are typically
concentrated portfolios
[291]
with relatively high fees focused
[293]
on trendy themes are launched
[295]
just after the peak of excitement
[298]
and often the peak of returns related
[300]
to popular investment themes.
[302]
They find that specialized ETFs
[303]
do not create value for
investors on average,
[306]
delivering a negative
four factor annual alpha
[309]
of about 6% on average in the
five years after inception
[312]
of the fund.
[313]
Before the fund inception,
[314]
the authors find that the indexes used
[316]
to create these specialized
index funds tend
[319]
to have very positive performance.
[321]
This makes sense economically
for the fund provider.
[325]
Create an index based on a popular theme
[327]
with recent high past returns,
[329]
build an ETF to track that index, profit.
[333]
Unfortunately for investors,
these specialized funds tend
[336]
to hold securities with
low expected returns.
[339]
Leading up to the thematic ETF launches,
[342]
the securities in the indexes tend
[343]
to have been increasing
in relative price measured
[346]
by book-to-market and media sentiment.
[348]
But pre-launch, these
indexes are not investible.
[351]
They are back tested conceptions
[353]
that index manufacturers think will sell.
[355]
After the ETF launches
[357]
and the index becomes investible,
[358]
the stock valuations tend
to come back to Earth
[361]
and the media sentiment tends to decline.
[364]
Thematic indexes also
tend to contain stocks
[366]
with more positive skewness,
[368]
which would be appealing for investors
[369]
who have a preference
for lottery-like payoffs.
[372]
They find that within specialized ETFs,
[375]
the ones holding stocks
[376]
with the highest pre-launch returns
[378]
and media sentiment performed
the worst post-launch,
[381]
further supporting the
idea that the issuance
[383]
of specialized ETFs occurs near the peak
[386]
of valuation of the underlying securities.
[388]
Now, surely, investors in these funds
[390]
don't think they are going
to lose money by investing.
[393]
The authors use analyst
forecast as a proxy
[396]
for investor expectations.
[397]
They find that portfolios
[399]
of specialized stock display
[400]
significantly higher
long-term growth forecasts
[403]
in the period leading
up to the ETF launch.
[405]
After launch, the stocks
experience a downward revision
[408]
in growth expectations.
[410]
The forecast errors, measured
by the realized earnings
[413]
per share minus the analysts'
long-term growth forecast
[416]
at the time of launch for specialized ETFs
[418]
becomes significantly negative
and economically large.
[421]
This is consistent with strong
over optimism in expectations
[425]
around the time of launch.
[426]
These data support the hypothesis
[428]
that the providers of
thematic ETFs launch products
[431]
based on themes where investors
hold optimistic beliefs.
[435]
Thematic indexes on which
thematic ETFs are based consist
[438]
of high-priced stocks with
strong media sentiment
[441]
and optimistic analysts' forecasts,
[443]
all of which tend to
decline around the time
[445]
that the ETF tracking the index launches
[448]
and investors can actually
start putting their money in.
[450]
A question worth exploring further
[452]
is what risk factors thematic
ETFs offer exposure to?
[456]
There are well-documented risk factors
[458]
that explain the majority
of differences in returns
[460]
between diversified portfolios.
[462]
How do thematic ETFs stack up?
[465]
In "Betting Against Quant:
[466]
Examining the Factor Exposures
of Thematic Indexes,"
[469]
David Blitz digs into the question
[471]
using major index providers MSCI
[473]
and S&P's thematic indexes as a sample.
[476]
Blitz finds that thematic indexes tend
[479]
to have lots of idiosyncratic risks
[481]
measured by their volatility ratio.
[483]
That is risk that could
be diversified away.
[485]
And they tend to be tilted
toward small cap stocks
[487]
with high prices, weak profitability
[489]
and aggressive investment.
[491]
Empirically, this
combination of risk factors
[494]
is terrible, delivering the
worst risk-adjusted returns
[497]
of any factor-based portfolio sort.
[500]
Why would anyone invest in these funds?
[502]
Blitz offers some ideas.
[504]
Investors in these funds
may simply not believe
[506]
in factor models for expected returns.
[509]
For example, any investor who believes
[510]
that value investing is old news
[513]
would have less of an issue
holding small cap growth stocks.
[515]
In other words, these
investors may not believe
[518]
in asset pricing models.
[520]
The other possibility
[521]
is that these relatively
concentrated portfolios
[523]
are delivering positive
risk-adjusted returns or alpha.
[527]
Despite their terrible
expected return profiles,
[530]
thematic indexes may still be
delivering positive returns
[532]
that are unexplained by a factor model.
[535]
In reality, while thematic indexes
[537]
do have positive alphas in back tests,
[540]
live funds, once launched,
[542]
tend to have negative
alphas, as we saw earlier.
[546]
Another possibility is that
some investors may want
[548]
the lottery-like payoff
of specialized indexes
[551]
if they have high
conviction in a given theme.
[554]
Who cares about expected returns
[556]
if you're just making a thematic bet
[557]
that the market has
underpriced a certain theme?
[560]
As we would expect with
lottery-like payoffs,
[563]
these will be losing bets on average
[565]
but winners could be big.
[567]
Or maybe investors are
simply behaving irrationally,
[569]
chasing past returns in
attention-grabbing investment themes
[572]
as they tend to do.
[574]
This behavior is well documented
[575]
in 2007 and 2020 papers by
Brad Barber and Terrance Odean,
[580]
among many others.
[581]
A final possibility is there
are non-financial benefits
[584]
to owning the stocks in thematic indexes.
[587]
Things like a need to belong,
[589]
feeling good about the investment
[590]
or satisfying a need for thrill seeking
[593]
are all documented reasons
[594]
that people choose to own assets separate
[596]
from their financial benefits.
[598]
I can definitely see this
[599]
for things like clean energy
[601]
or even ARKK where
investors form a community
[604]
around their investment.
[606]
The title of this paper,
[607]
"Betting Against Quant:
[608]
Examining the Factor
Exposures of Thematic Indexes"
[611]
is alluding to the fact
[612]
that quantitative investors
[614]
are often investing in stocks
[615]
with low prices, robust profitability
[617]
and conservative investment.
[619]
Investing this way can
be done systematically
[621]
at a low cost and has theoretically sound
[623]
and empirically persistent
positive expected returns
[626]
in excess of the market.
[627]
To build these portfolios,
[629]
quant investors need someone else
[631]
to overweight the small cap growth,
[632]
low profitability, high investment stocks
[635]
since all stocks in the
market needs to be owned.
[637]
It seems like thematic investors
[639]
may be taking this junk off the hands
[641]
of quant investors.
[643]
Regardless of their financial performance,
[646]
another big hurdle for investors
[647]
is their own behavior.
[649]
In Morningstar's 2021 Mind the Gap report,
[652]
which compares the return of funds
[654]
to the return of the
investors in those funds,
[656]
which are affected by the
timing of their cash flows,
[658]
one of the biggest behavior gaps
[660]
is for sector funds.
[662]
Morningstar doesn't categorize
thematic funds in this report
[665]
but I would be comfortable betting
[666]
that the behavior gap is even larger.
[669]
This was particularly problematic
[670]
for Cathie Wood's flagship ARKK ARKK fund
[674]
as documented by Morningstar.
[676]
At the end of November 2021,
[678]
ARKK Innovation had delivered
[680]
a staggering five-year
compound return of 41.3%
[684]
but the average ARKK investor
had earned only 9.9%,
[688]
not to mention the absolute
collapse since then.
[691]
I tried to warn you guys.
[692]
Fund companies have a huge incentive
[694]
to create and market products
[696]
that cater to attention-grabbing
investment themes.
[699]
The fees on thematic ETFs are a lot higher
[701]
than on broad market ETFs.
[703]
Take Cathie Wood's ARKK Innovation ETF,
[705]
which has a few of 0.75%
[707]
or the Horizons Marijuana
Life Sciences Index ETF
[710]
with a management fee of 0.75%
[713]
and a management expense ratio,
[714]
including other costs of 0.85%.
[717]
This is way more expensive
than a broad market ETF
[720]
or even a small cap value or value ETF.
[722]
It is good business for ETF providers
[724]
to ride these trends
[726]
but for investors, jumping
on thematic bandwagons
[729]
is one of the worst approaches
[730]
to investing money, both in the short run,
[733]
unless you get lucky,
[734]
and definitely in the long run.
[736]
Thanks for watching.
[737]
I'm Ben Felix, portfolio
manager at PWL Capital.
[740]
If you enjoyed this video,
[741]
please share it with someone
[742]
who you think it could
benefit from the information.
[744]
We also discussed this
topic in episode 185
[748]
of the Rational Reminder Podcast.
[749]
(upbeat music)
Most Recent Videos:
You can go back to the homepage right here: Homepage