Can You Really Game Index Funds?

http://www.bloombergview.com/articles/2015-07-07/can-you-really-game-index-funds-

by zweiterlinde • 11 years ago

115 48 11 years ago

bloombergview.com

kqr2• 11 years ago

Reference to hn thread discussing the original article A Profitable and Legal Way to Game the Stock Market

https://news.ycombinator.com/item?id=9844686

solve• 11 years ago

A finance writer that actually knows what he's talking about, and it's here on HN. This is nice.

arbitrage314• 11 years ago

Agreed! Very nice and rare to see on HN! :)

KingMob3• 11 years ago

Ehh, not quite. The author is indeed correct about the market-makers providing liquidity to everyone who wants to purchase on the day a company is added to an index.

But saying "index funds free-ride on the work done by active investors" and then following with "no one thinks that active managers should be able to charge for their services, is a world that will spend too little time and effort on allocating capital to the right businesses" is FUD.

The value of the market represents the sum total opinion of everyone in it (plus noise), not just the managers of mutual funds losing business to index funds. Frankly, it sounds like the griping of someone telling fund managers that they deserve their fees, but the supposed loss from using index funds described in the original article (~.2%) is still dwarfed by the increased fees of actively managed funds.

Most index fund expenses are around .1-.2%, while most active funds come in at a whole 1-2%. To justify the cost of an actively managed fund, a manager has to not just beat the market, but trounce it. Very, very few can do so for any length of time, and they know it, which is why articles trying to convince people of the virtue of active fund management are constantly written. Unless your manager is as good as Buffett, buy an index fund.

The math is simple, but there's many fund managers out there trying to convince you otherwise.

URSpider942• 11 years ago

I didn't necessarily take him to be saying that. He is absolutely right that a market that is, quite literally, 100% passive would just sit there and not do anything -- it would just grow as money comes in, but there wouldn't be any relative movement of one share against another.

However, we are in no danger of running out of active traders, so there's no need for anyone to run out and sell their indexed investments to save the free market ...

tptacek2• 11 years ago

Levine does not think you should invest in actively-managed funds.

The little coda about active management makes more sense if you read him religiously, because this is a schtick of his. Passive management helps most investors. But the market as an entity benefits from active management, because active management makes prices more accurate. This despite the fact that for the most part, contributing to the accuracy of prices comes at the expense of the actively-managed funds.

So without active management, the passive funds would perform more poorly, because their prices wouldn't benefit from the corrections of people trading into them to profit from mispricing.

mikeyouse• 11 years ago

There should be a bot that posts everything that Matt Levine writes to HN -- It's such great quality writing from someone deeply knowledgeable in finance.

Another good article where he discusses the Goldman software developer who 'stole' company code:

http://www.bloombergview.com/articles/2015-07-07/goldman-cod...

cma7• 11 years ago

Order-handling companies pay for "dumb" flow. Vanguard can reduce their outright trading costs to negative by being as dumb about it as possible, and then use these negative costs to artificially lower their reported fees.

Just because Vanguard claims to be smart about it, doesn't mean necessarily they actually are incentivized to be smart about it or actually are in practice. People can still judge them by how close they track the index, but that is reported separately from fees, which are all a lot of current and future retirees look at after having fees fees fees drilled into their heads. And the indexes themselves take a hit, so you need to adjust for that with a much more complicated measure.

They can effectively launder bad (or even good) tracking of the index into lower reported fees, by letting the order handlers profit on the inanity (and kickback via order-flow payments), allowing the fund managers to give themselves higher compensation without commiserate alarming fees.

To what extent, if any, do they actually do this? Do they report their income from paid order-flow in the fund prospectuses? Do they break it out by the managed funds, vs retail flow from their clients? Do they get major concessions to their retail trading costs in tacit exchange for being dumb with their etfs?

iaw• 11 years ago

Could you elaborate or provide more reference materials on "dumb" flow?

cma7• 11 years ago

http://www.reuters.com/article/2010/12/17/us-markets-dumb-mo...

It is basically order flow that often blindly takes liquidity. Traders pay to get it, execute it prop at the best price (though they also can dump straight to the market without taking it prop if it isn't behaving dumb enough). Then they resell the position on the market at a more deliberate pace, providing liquidity (which lets them capture the spread, and lets them earn kickbacks from exchanges that pay for liquidity). Or they match the positions against future dumb flow coming in on the other side.

jsprogrammer2• 11 years ago

So, who loses money in your situation? The person who doesn't realize this is happening?

cma7• 11 years ago

Shareholders of the ETFs. Again, I'm not saying this specifically actually happens in the case of Vanguard, but all kinds of other conflicts of interest like this happen all the time, and reporters should be ever skeptical, and not just report he-said she-said, "oh Vanguard says they got this covered."

snowmaker• 11 years ago

Is there any evidence that Vanguard gets kick-backs in return for their dumb order flow?

I would think that would be a HUGE scandal if it were true and ever came out.

cma7• 11 years ago

I haven't offered any, and it would most likely be illegal if it was explicitly going on. And there is likely a wall between the different trading desks (though often the physical embodiment portion of this is literally a cubicle wall the employees can hear each other over). But the orderflow compensation doesn't need to cross over into the retail desk if the ETF itself has enough trading volume to mask some kickback without appearing too egregious.

But price fixing is also illegal--nevertheless, two gas stations across the street at a profitable intersection can engage in it solely through price signal tit-for-tat[1]. This effectively masks intentionality.

Much more fantastical and speculative: machine learning algorithms at both firms could now, or in the future, arrive at this cooperative strategy, even with the only communication being through price signals. Without any human ever even knowingly giving the explicit go-ahead.

[1] https://en.wikipedia.org/wiki/Prisoner's_dilemma#The_iterate...

kasey_junk2• 11 years ago

Vanguard Brokerage Services disclosure about order flow payment can be found at:

http://vrs.vista-one-solutions.com/reports/1-6/vang/

Short answer is that Vanguard Brokerage Services did not receive compensation for equities order flow in 1QTR 2015, but did receive compensation for options order flow.

I have no idea what the laws are surrounding their index funds or how those index funds are allowed to/actually do interact with Vanguard Brokerage Services.

murbard22• 11 years ago

The company managing the Vanguard fund is owned by the fund holders themselves. It's a closed system and the incentives are aligned.

cma7• 11 years ago

There is a principal agent problem between managers and shareholders. Managers get paid out of shareholders' pockets. If the shareholders are over-focused on fees as a metric, which they often are, managers have some potential tools (which I went into) to launder part of the fees into poor fund performance. I'm just throwing it out there, I'm not saying it is actually going on.

---

As for proxy voters choosing the management, proxy vote research usually doesn't make sense. You see strategies amongst large hedgefunds like this: keep their valuable proxy vote research private, make a vote based on the most predictable-to-them-but-not-to-you outcome based on the research (positive or negative for the company, doesn't matter unless it will be picked up by others immediately, even then they can do things like exiting their position through an obscure hedge), and then hold or sell their shares based on the overall vote outcome and its implications in the research.

Researching and making proxy votes out of naive benign interest of the company is often just doing altruistic work for a greater collective, something markets frown on and usually punish.

arielweisberg2• 11 years ago

If it were 20+ basis points a year it should show up in the returns and as a failure to track the index. I am not an expert, but that isn't what I see eyeballing a chart of VFIAX over 35 years. It doesn't track perfectly by an amount that does matter, but not .20 basis points a year.

Also by this logic total market funds should outperform other indexes by a healthy amount over time. Also maybe not what we are seeing. Granted total market funds invest in something that is different from what other indexes track.

cma7• 11 years ago

No, the 20+ basis points is considering how the index itself underperforms, because stock prices get bid up just before they get added to the index and then drop back down as the liquidity crisis settles. The index, not just the etfs, take a hit. Vanguard claims to soften this by trading more deliberately and not buying or selling it all at the opening auction on the day a stock gets added or removed, respectively. So you would expect them to be able to do better than the actual index (until you add in all the other management costs/trading costs).

To an extent the amount the ETFs are effective at this lowers the amount the index underperforms, because they (the etfs) are themselves the driver of the liquidity crisis the index is getting subject to. So you would expect some sort of equilibrium, and the claim is therefore to be taken that this equilibrium settles down at of 20+ basis points.

URSpider942• 11 years ago

And yet, the index itself (not even the funds) reliably beats the overwhelming majority of active traders over almost any time window you care to look at.

arielweisberg2• 11 years ago

Thanks that is very enlightening as to how this front running manifests.

Still makes me wonder if that means I am going to get an extra 20 basis points for free out of VTSAX.

tptacek2• 11 years ago

Even if you don't care about the index fund "front-running" "scandal", the section starting at "The value of market-making is hard to see and easy to criticize" is critically important to understanding why the markets work the way they do.

As always, Levine is fucking fantastic.

6stringmerc• 11 years ago

You know what's easier to criticize? Manipulating LIBOR and nobody going to jail. The downstream effects of instruments pegged to LIBOR is staggering, well, would be if the industry / reglatory agencies actually did anything of merit.

Remember, the main rationalization for Bernie Madoff's unbelievably consistent returns was that he was front-running, and fund after fund after fund after fund lined up to give him money. Besides, front-running isn't really where the big money is anyway. Insider trading is way, way more profitable from an individual standpoint.

anonu2• 11 years ago

This is called the "index rebalancing" trading strategy. Prop desks and hedge funds have known about it for decades. A lot of money is passively benchmarked to many popular indices provided by the likes of S&P, DJ, Nasdaq, etc... One reason people invest in funds that track these indices is because they believe the index provider is a good benchmark for whatever its tracking. For example, the S&P 500 tracks the 500 largest US names. The Nasdaq 100 tracks the 100 biggest (mostly tech-related) names that are Nasdaq-listed. etc... In addition to being a good benchmark, a set of rules (here are S&Ps: https://us.spindices.com/documents/methodologies/methodology...) are published by the index provider that govern how stocks are added and removed to the index. Understanding these rules allows arbitrageurs (aka market-makers) to predict when names are moving before they are announced by the index provider. Since a fair amount of capital is already tracking these indices, the passive indexer will be required to buy/sell the names in the index in the right proportion so as to be properly benchmarked.

Another interesting point is that the Volcker Rule has more or less caused a massive shift of this type of strategy away from US investment banks and into hedge funds. I don't have real data on this - just my observations.

tempestn2• 11 years ago

From what I can see, this article is on point, but is missing an important factor: the risk these "front runners" take. As soon as the announcement is made that a company is joining the index, it's public knowledge. In theory, the expected increase, minus a risk premium, should be priced in immediately. There will likely still be money to be made over the following days until the addition is complete, but it's far from guaranteed, and comes at the expense of reduced diversification. (Which I suppose is another way to say that you're getting paid for providing liquidity, as the article says.) Just because AA went up X% over the 4 days, or whatever, before it joined the index, doesn't mean the next stock will. Perhaps its jump will be overestimated by the HFTs, and retail investors trying to get in in the days following the announcement will end up losing money. Probably not, but it's certainly a significant possibility. So if a person wanted to pursue this active strategy, they would need to manage their risk appropriately. It's not necessarily a bad idea if you enjoy spending your time on that kind of thing, although personally I'd rather index (with a moderate small/value tilt).

anonu2• 11 years ago

most of the juice is in predicting the move before the public announcement

tempestn2• 11 years ago

Likely true. Also not without risk of course, since the chance of a stock getting added should also be priced in. If you're better than the "market" at predicting these things, you'll likely do well. I don't expect I am.

kasey_junk2• 11 years ago

More commentary from an ex-index arb trader:

http://kiddynamitesworld.com/where-bloomberg-discovers-that-...

meeper16• 11 years ago

How about a historical back-test? Should be simple enough.

danieltillett2• 11 years ago

Are there any index funds that track the entire market? Wouldn’t investing in all companies solve the problem of index tracking?

p1mrx• 11 years ago

Yes, it's called a Total Stock Market Index Fund, and they are quite popular. There's really no reason to invest in the S&P500 anymore.

danieltillett2• 11 years ago

That makes sense. I am surprised why anyone would want to choose S&P500 in this case.

atinoda-kestrel• 11 years ago

Are there any index funds that track the entire market? Wouldn’t investing in all companies solve the problem of index tracking?

Yes, there are funds that try to match the entire market. Vanguard runs a couple "total market" indices (VT and VTI, for example), and there are more.

But that's not to say that they actually invest in the entire market. They don't; they use sampling to try to match the performance of the thing they index. (Vanguard, at least, gets very very close to this...)

Disclosure: I am invested in in VTI.

jackgavigan• 11 years ago

Matt Levine stands out amongst journalists and commentators as someone who actually knows what he's writing about because he's been there, done it, and now wears the t-shirt when he's changing the oil on his car.

Can You Really Game Index Funds?

Related Stories