[digitalatlas]

This article was posted on Yahoo, and among other topics, they discussed how Burry thinks there's a bubble created by the increasing popularity of indexing. I hadn't really thought about this before, but it seems he suggests that (if I'm understanding correctly) that the prices are not reflecting the true value of the underlying equities, but indexes have risen because people keep putting money into it. I can see there being some logic to that. What do you think?

https://finance.yahoo.com/news/big-short-michael-burry-explains-104146627.html

 

[Topo]

I am not sure that funds flowing into index funds is any different than funds flowing into single stocks in the market. If the stock market is overvalued or is in a bubble (doubtful), it is not because investors are using index funds. It would be no different than buying single stocks, in aggregate. In a way, inflows to broad-based index funds are neutral, because they are proportional to the weightings of each stock in the market.

 

[digitalatlas]

It is the same in principle, you're right. But the size of the cohort of index investors, the historically unprecedented ease of purchasing an index, could potentially influence how much money is flowing into the market, beyond what influence (in the past) a smaller, more knowledgeable group of professional investors may have bought, based on fundamental analysis, could it not?

I suppose one could try to calculate or estimate what point of intersection that may be, but it would probably be hard with all the variables involved. But that point should exist.

My background is actually in the sciences, and as an example, when we look at nanoparticles on human exposure leading to disease, let's say you take a carbon atom that's not a nanoparticle. It might cause one health effect. Take the same carbon source, but break it up into nanoparticles, and even though it's the same stuff atomically, it behaves differently and can have a different health outcome.

 

[Topo]

It is the same in principle, you're right. But the size of the cohort of index investors, the historically unprecedented ease of purchasing an index, could potentially influence how much money is flowing into the market, beyond what influence (in the past) a smaller, more knowledgeable group of professional investors may have bought, based on fundamental analysis, could it not?

It may be true that there is more money flowing into stocks than the past, but I think this is because money is cheap, we are awash in debt, and there is a perception that stocks are less risky if held for the long term. The intermediary/vehicle probably would not make a major difference. For example, once I decide to invest X amount into stocks, whether I buy a few individual stocks, an index fund, or give it to a fund manager to pick stocks for me, would not be materially different: X amount is flowing into stocks. The "market" is the aggregate of all the participants, whoever they may be or whatever vehicle they use.

 

[OptsyEagle]

1st of all. There were a lot more factors to the previous credit crises (unethical mortgage brokers, unethical rating agencies, unethical bankers, unethical brokers and insurance companies that had no idea how to price credit risk...and borrowers who thought they could buy a $750,000 house earning $15,000 per year of annual income) then just the pricing of CDOs, as this individual seems to suggest.

2ndly, the credit markets, where CDOs existed, were a little more difficult for the average investor and even the more sophisticated, to go to work and analyse. Contrast that to the stock market, where just about everyone can access corporate news and financial reports, etc., and you will see a drastically different market for research.

With all that said, he is correct that index investing can distort the proper price of a security within that index, one needs to keep in mind that there are literally 100s of millions of investors attempting to take advantage of any distortion on any security at any time. Since the analysis of most index securities is much easier then some of the convoluted debt securities that existed during the credit crises, I have full confidence that those investors are quickly correcting any distortions, that index investing might create, long before they get anywhere detrimental.

I would worry when you start to see index investing becoming more then 25% of equity investing. Where 100% of investing incorporates all individual investors as well as institutional, like pension funds, mutual funds, hedge funds, etc. The last time I looked we are still under 10%. I doubt it will get up to 25%, because of our basic greed instinct, within the human mind. Remember, that an index fund will probably be outperformed by 40% of all mutual funds and maybe even more of all individual investors...AT ANY GIVEN TIME (not necessarily long term). Humans hate seeing all these investments outperforming their own, at all given times, so they tend to have this constant aversion to indexing, right up until they start seeing their own results from those alternatives.

Anyway, we are a long way away from the point that these small distortions will create a serious drag and even when that happens, it will be a small percentage drag, NOT a credit crises type implosion.

 

[Rusty O'Toole]

There are a few angles on this. One is that the stock market as a whole has been driven for the last 10 years by massive money printing by central banks, and their zero interest rate policies which forced investors to buy stocks in hope of getting some kind of return, as opposed to bonds which have practically no return, and in some cases negative returns.

Another is that indexing means every stock on the list gets bought willy nilly. So a lot of bad and mediocre companies get bid up for no reason except that they are part of an index.

There is another way to look at it, and that is, that all the ETFs, mutual funds, hedge funds etc are the market. They account for the majority of stocks bought, sold and owned. This is why they fail to beat the market - in aggregate they are the market.

The only way for the boom to crash is if the central banks stop printing money and interest rates rise to more normal levels. But that cannot be allowed to happen because it would crash the world economy and bankrupt every government and large company. All are in debt and all depend on low interest rates to survive. If interest rates went to 5 or 6 percent their payments would double or triple and they don't have the money to pay it.

So, I don't see the investment environment changing any time soon. I agree that fundamentally it is a **** show that would collapse overnight if rational policies were put in place. And that is why rational policies won't be put in place. It's too late to even think about doing so.

 

[james4beach]

Here's another article on it: The Big Short’s Michael Burry Explains Why Index Funds Are Like Subprime CDOs

Haven't read it yet, I'll read it when I get a break today. I suspect he's going to say that you pick up a lot of junk in the index, and can do better by stock picking. Basically an argument for active management.

 

[Topo]

I don't think buying the index would distort the prices of the constituent stocks. The prices would go up and down, keeping the relative weights constant. If one were to go into the market and buy a stock individually, and in the process bid the price up or down, then the price would be distorted relative to other stocks.

Buying an index is neutral with regards to the weightings of the underlyings. So if I buy SPY, the relative proportions of MSFT, AAPL and AMZN will not changes, but the price of everything in SPY could go up or down, depending on how I bid.

 

[OptsyEagle]

The weightings in the index do change as the prices of the underlying stocks change. The exceptions would be equal weight ETFs and the like.

There is a distortion, but nothing that might create a credit crises like implosion. There are too many non-index investors correcting those distortions, everyday, before they get too far out of line.

Now if one could accurately calculate and identify what is a distortion and what is not, then they could take advantage of such a thing. Since most humans will be right on some and wrong on others and will pay fees each time they trade, the index investor will still win out, in my opinion. Even with those little distortions. In other words, it would be better to eliminate the huge distortion that seems to come from being a human being, which is what one can almost do with index investing.

I imagine the author of that article, with all the fanfare he has received since 2009, including a movie about his endeavours, must have quite the distorted opinion of himself by now. One of these days, that is going to become very costly to him and his investors. He should probably think about index investing.

 

[andrewf]

Sounds like a lot of FUD. The best argument that critics have is that index investing in illiquid underlying assets (such as small caps or foreign small caps) could lead to illiquidity in those index funds. Sure. But most assets are invested in liquid assets such as the S&P 500. I fail to see how a period of illiquidity in small-cap funds will cause a collapse. The underlying companies won't go bankrupt due to a short spell of panic selling.

The risk with index investing is that assets get mispriced relative to each other. But that has a natural stabilizer in the form of active investors. I don't think passive investing is at any risk of eating the world, especially for long term investors who would be making up a small share of trading activity besides.

 

[Topo]

The weightings in the index do change as the prices of the underlying stocks change. The exceptions would be equal weight ETFs and the like.

The weightings change, but not because of index investors buying the index. They change because other investors are buying and selling subsets of the market (or because of buybacks, secondary offerings, etc). This would only apply to a cap-weighted index (such as the S&P500).

 

[james4beach]

Here's another article on it: The Big Short’s Michael Burry Explains Why Index Funds Are Like Subprime CDOs

I think some of his points are valid, but he's being overly dramatic. The articles are implying some kind of disaster outcome like the total wipe-out in subprime CDOs. That is simply not going to happen with plain vanilla stock index ETFs. With exotic ETFs, yes it will - avoid those.

I think a better phrasing of Burry's argument is: "You're going to get more volatility, and sharper drawdowns, in index funds due to price distortions that are building up from indexing" (my words).

And I agree with this. I think in then next crash or bear market, investors who bought into the idea of indexing without sufficiently appreciating risk will experience sharp drops that they never anticipated. I personally expect another bear market or crash in stocks could take the indexes down 50% to 70%. Since each crash is different and always surprises investors, this could either turn out to be sharper (larger max drawdown) than in the past, or much faster than people anticipate, causing panic.

For example, when I run into a guy in the gym locker room who tells me he's new to investing and is putting all his savings into VGRO (as I did a couple days ago), that's the kind of person who is going to get destroyed in this scenario, and will probably panic sell due to the incredibly sharp % drawdown. These are people who have no idea what they could be in for. When this guy experiences a 60% drawdown in VGRO, he's going to lose his mind. He'll be thinking "this thing was supposed to be diversified! this kind of decline should be impossible! indexing must be broken!"

I think Burry could turn out to be correct that due to major price distortions from indexing, that the index could have these severe drops, whereas stocks from "outside" the index, or a fundamentals-based portfolio, could perform better -- relatively speaking. Another way to describe this would be that we could see indexes like SPY and XIC drop a much larger % than actively managed portfolios that aren't relying on indexing.

If that scenario unfolded, it would make indexing look "broken" for a period of years (due to underperformance). That could be 5 - 10 years ... which is a very long time for investors ... where the index gives the worst results. However, this would not mean indexing is broken. It's just the natural result of very large price swings (overvaluation > mean reversion > undershoot > back to normal)

But that's the worst case outcome that I see from this argument. The indexes won't blow up and become useless. Rather, the investors who rely on indexes get more severe drawdowns, severe volatility, and then lingering underperformance. Volatility is in fact quite dangerous, something people don't believe, but due to human psychology and real-life time horizons, I strongly believe that volatility is a very hazardous thing.

He does raise another point though, which is about the structured ETFs which use derivatives techniques to sample and replicate indices, and how hard it will be to unwind or adjust those positions in market turmoil. On this one I agree 100% and humble_pie has been warning about this for some time. The more complex and ambitious ETFs, which supposedly hold huge numbers of stocks, take shortcuts and use alternative techniques to synthetically track indices.

Those kinds of ETFs, of which are there are a ton (including IVV) will probably then underperform their respective indexes.

This is a big reason I continue to like XIU, my favourite Canadian ETF,. This is a true "plain vanilla" ETF, and it only holds 60 stocks. That's important because these are the 60 most liquid stocks in Canada... a tremendously good design choice. It is totally feasible to effectively track and manage funds in & out of 60 of the most liquid large caps. At the danger of speaking for humble_pie, I think she and I agree that it is not feasible to accurately track and manage fund flows for one of these ETFs like VT which supposedly holds 8,202 stocks from around the world, or XAW with 8,653 supposed holdings.

~~ Summary ~~

I think indexing is still fine, except we may get severe volatility and huge drawdowns beyond expectations of most investors, due to price distortions we've had in recent years as described by Burry. Personally I'm expecting equity index declines (max drawdown) of 50% to 70%. Volatility is dangerous, so I think it's an important warning. Particularly bad trouble could hit the more exotic ETFs, and structured funds which use derivatives to estimate and mimic various things.

To mitigate these risks, stick to the tried-and-true plain vanilla ETFs, the simpler the better, and be prepared for the possibility of very large % drawdowns in excess of the 2001 and 2008 bear markets.

You might want to avoid the global ETFs which claim to hold huge numbers of foreign stocks. I avoid those. I think they will likely underperform due to some derivative and synthetic replication blowups. I prefer XIU and ZSP (plain vanilla) over XAW, for example.

 

[john.cray]

He does raise another point though, which is about the structured ETFs which use derivatives techniques to sample and replicate indices, and how hard it will be to unwind or adjust those positions in market turmoil. On this one I agree 100% and humble_pie has been warning about this for some time. The more complex and ambitious ETFs, which supposedly hold huge numbers of stocks, take shortcuts and use alternative techniques to synthetically track indices.

That's one thing that caught my attention in his comments and I immediately remembered what humble_pie has been warning us for some time. Thanks for the reminder.

One question to you and her: How can one tell if an ETF uses derivatives to track an index versus one that doesn't, i.e. strictly holds the underlying stocks directly?
Is looking at the "holdings" section of the ETF's page enough or is there another way?

Is the sheer number of underlying stocks a good indication of derivatives? For example I hold VIU which has 3,722 holdings. In the factsheet it reads:

Currently, this Vanguard ETF seeks to track the FTSE Developed All Cap ex North America Index (or any successor thereto). It invests directly or indirectly primarily in large-,mid-and small capitalization stocks of companies located in developed markets, excluding the U.S. and Canada.

Is the *indirectly* part referring to using derivatives or simply other ETFs?

 

[andrewf]

Can't one look at the creation unit? Ie, the basket of securities that a authorized participant needs to provide in exchange for a set number of units of the ETF? That should be a good guide to what the fund holds, as it also what the fund needs to provide in the case of redemptions. An ETF will run into problems during fund outflows if it does not hold this basket.

 

[OnlyMyOpinion]

Just look up the fund prospectus and perhaps some annual reports on SEDAR or the fund site. This should be done before you buy into any fund anyway.

 

[james4beach]

One question to you and her: How can one tell if an ETF uses derivatives to track an index versus one that doesn't, i.e. strictly holds the underlying stocks directly?
Is looking at the "holdings" section of the ETF's page enough or is there another way?

I don't know, exactly. humble_pie suspects that regulations don't require complete disclosure on this. She may be right; I don't know enough about securities regulations.

I look at the audited financial statements and take that literally. I see enough things in there to raise my suspicions, as described below.

The holdings web page is not sufficient. You do actually have to look at financial statements. For example, pulling up IVV (the US ishares page, click Annual Report) and browse to page 19, Schedule of Investments, you start to see interesting things:

You will see under several sector categories that they list some stocks, but not all the stocks of the S&P 500. There are many placeholders that say "Other securities", for pretty substantial %s of the ETF. Those are probably derivatives like futures, or who knows what else. Some of them appear like that because the actual stocks are out on loan (securities lending). Many stocks are missing.

There is also a section showing futures contracts. They hold S&P 500 E-mini contracts, so to some extent, IVV is mimicking the S&P 500 using index futures.

Determining the extent of the futures exposure (in lieu of real stocks) is tough to assess. It's hard to understand because the value of the futures usually appears to be tiny. What people forget, though, is that futures are very highly leveraged. I don't remember the exact numbers, but it's something like, $10 million in futures is equivalent to $300 million or more in notional exposure.

I think IVV and the other iShares core funds are one example of derivatives infiltrating what should be a simple, plain ETF. And remember that Canada's XUS just is a wrapper around IVV, so you have all the same futures fun.

It's very tricky to dissect actually. I usually look out for any sign of derivatives holdings, and then avoid them. I would not invest in IVV or XUS.

How about Canadian ETFs though? Ugh, I'm not sure. I don't know if disclosure is as good with ours. Let me take a look at XAW. Again, look at the Annual Financial Statements linked at the bottom of the iShares page.

This is one of these "structured products", a wrapper of ETFs. Personally I do not like all the layers of indirection in these things (wrapping and sub-holdings).

XAW holds a bunch of other ETFs. It holds XUS which in turn holds IVV, which as discussed above, includes futures contracts (index derivative in lieu of real, individual stocks). So yes, XAW contains some derivatives and index replication techniques.

XAW's next largest position behind XUS (IVV) is XEF. According to this financial statement, it holds a monstrous list of securities in many countries. I don't see any placeholders or explicit derivatives, but then again, this is written according to Canada's securities laws (maybe as humble_pie says, they don't have to disclose replication tools or derivatives).

XEF really shows a ton of underlying securities. I just have the nagging suspicion that they can't really manage all those positions in all these securities, many of which will have limited liquidity. How could they? It just doesn't seem feasible. Go look at the list yourself. Then look up some of the stocks on European and Asian exchanges ... take a look at how little volume some of them. Random example: accesso Technology Group in the UK. This traded just 7,000 shares in the day in total.

Maybe I'm overly cynical but this just doesn't sound right to me. What happens when $50 million flows into XEF in a single day. Does iShares go out and buy another 3 shares or whatever (odd lot) in a highly illiquid security that only trades 7,000 shares a day?

Think of how big the bid/ask spread must be on something that illiquid. And XEF is moving $ in and out of that, daily? Really?

(a) if they're really doing that, this isn't particularly efficient, and there will absolutely be friction and big losses in market turmoil where lots of $ is pulled out of XEF. This would cause high volume selling on already illiquid securities in a depressed market, causing huge friction and inefficiency

(b) Or, as humble_pie suspects, are they replicating it somehow, and Canada's securities laws doesn't show us the nitty gritty details? This would be more efficient and potentially more desirable, but also gets into games.

At face value, based on what's written here, I don't see any clear signs there is sampling replication, or other derivatives. I don't know if it would show up, though. I have my suspicions. I don't know which I would prefer, (a) or (b) ! Neither is good.

Is the sheer number of underlying stocks a good indication of derivatives? For example I hold VIU which has 3,722 holdings. In the factsheet it reads:
Is the *indirectly* part referring to using derivatives or simply other ETFs?

I think that's what the language is referring to. But I don't think so clear, just based on the number of holdings.

Let me see if I can see what this "indirectly" refers to... pulling up financial statement...

There's a statement buried on page 308 which may shed some light on this:

Additionally, the Funds may use futures contracts to maintain full exposure to the stock market, maintain liquidity and minimize transaction costs. Funds may purchase futures contracts to immediately invest incoming cash in the market, or sell futures in response to cash outflows, thereby simulating a fully invested position in the underlying index while maintaining a cash balance for liquidity.

I do see some evidence of futures existing in many of these funds, but the accounting of it really isn't clear to me. To be honest, I can't tell from these financial statements what's going on.

Here's my guess. VIU, like XEF, shows an absolutely massive listing of individual securities. Tiny holdings (like 233 shares) in some totally illiquid foreign stock. How the heck can this seriously be managed, given that the various shares trade in different time zones, out of sync with Toronto trading. Apparently they are buying 1 share or selling 2 shares here or there in illiquid foreign markets, matching fund inflows/outflows of millions of $ in Toronto. Really?

Maybe that's why the futures are used, replicating index exposure, and then perhaps they wait and group the securities buy/sells into larger blocks (which would be more efficient).

Sorry... no clear answers. With those US funds, it's pretty clear they hold futures, and are doing something else which results in missing securities (could be replication, estimates, securities lending, or futures in lieu of stocks).

With the Canadian funds, it's less clear to me, but the financial statements seem to pretty explicitly state derivatives are used for "simulating a fully invested position".

 

[james4beach]

After delving into those examples, I think I now have a better idea of Burry's concerns, and I buy his argument more than I did before.

Those massive listings of tiny stock holdings in thousands of foreign stocks (VIU, XEF, XAW, etc) do really make me scratch my head. Clearly, there will be liquidity concerns here -- which will only appear during market turmoil. None of this comes up during good times.

Most of these ETFs did not exist back in 2008. They have never lived through that kind of market stress. Very important to think about!

There are plenty of signs of index replication using futures to maintain placeholder positions until catch-up stock trades can be made. As Burry says, this really does have a similarity to those mortgage things and structured products of the past crisis:

The index futures are single securities which (in a very abstract way) represent the underlying. They become a thing of their own, with all these funds trading the index futures among themselves.

Those futures get resolved, matched to real stocks at some point. Under normal conditions that is just fine. But under market stress, what happens to all those thinly traded foreign stocks (or thinly traded small and mid caps) with huge bid/ask spreads?

It's a mis-match of markets and liquidity. Very liquid index futures, which will probably diverge significantly from the actual underlying equities, during market crashes.

That's a parallel to baskets of mortgages (indices) which looked fine, until liquidity dried up in the underlying, and then all the structures fell apart. These ETF structures aren't quite as bad, because they do hold underlying equities, but the replicated parts are concerning.

And again, this wasn't tested in 2008. The wrapper-of-wrapper-of-wrapper-with-futures-placeholders things are totally new.

The implication of all this is that due to the friction that results from illiquid underlying, huge bid/ask spreads, and crashing futures trading against illiquid stocks... is that ETFs that contain too many of these replication derivatives may experience some losses (permanently) due to liquidity and tracking issues, that they will not recover when the market bounces back. It's just friction and liquidity related losses.

That will result in underperformance. The index on paper will look fine, and the ETF will do worse. How much worse, no idea!

 

[Topo]

The ETF's NAV could for a short time period, due to severe stress or a flash crash, etc, deviate from its true value, but at some point has to reflect its holdings. The same factors would also affect mutual funds, including active ones, so I don't understand why Burry is singling out passive investments. We could have a bubble of the same magnitude if all of us invested the same amount of money in active mutual funds.

There are decades of experience with mutual funds (including the 1987 crash, flash crashes, etc) and so far in most cases (index or active), there have not been any major problems. ETFs are basically nothing more than a tradeable mutual fund.

The investor could take the possibility of NAV deviations from market price into account and avoid trading ETFs or mutual funds during tumultuous markets. Of course, these times could also provide opportunities for some.

When the Greece problems surfaced a few years ago, the stock market closed for a while and only ETFs tracking Greek equities were trading. When the Greek stock market opened, the stocks traded close to their pricing in the ETF. It seems that freely trading ETFs provide price discovery similar to a functioning market.

 

[james4beach]

There are decades of experience with mutual funds (including the 1987 crash, flash crashes, etc) and so far in most cases (index or active), there have not been any major problems. ETFs are basically nothing more than a tradeable mutual fund.

But I think index futures for providing (artificial) liquidity is the new element. Traditional ETFs did not use them; ETFs are supposed to be baskets of stocks, redeemable and creatable for underlying stocks.

But now, index futures are allowing ETFs to track baskets of things that are normally not liquid and trackable, like smaller cap stocks, and foreign stocks.

Even if mutual funds used some futures positions before, they are settled once a day and don't have the same dynamics. An ETF trades continuously through the day, and traders react to the prices. So I really think this might be a brand new thing... futures providing artificial liquidity in real-time, continuously traded funds.

 

[Topo]

Security lending, it seems to me, is the Achilles Heal of these structures. It is common practice with most mutual funds and ETFs, and it introduces counterparty risk which will show up at exactly the wrong time. This is not unique to ETFs; brokerages lend stocks from personal margin accounts, which could be as much risky as when an ETF does this. This is not unique to passive funds though. Any XYZ held in a margin account has this risk.