I expected to see (1995) tagged to the end of this article.
I personally know 3 people who run their own money ( <5 million ) who do this as their sole form of income.
Having said that, this isn't exactly easy. You need to know
1) if a stock is going into the index
2) when its going into the index
3) how much the index will buy
4) how much the index buy will affect the price of the stock
the first 3 are trivial for some index funds, though most have rules that allow them some leeway here so there aren't always sure things.
The 4th is where you make your money.
And it isn't like there aren't other's doing this, the article makes it seem much easier than it actually is.
If you play the game theory through you'd realize that if you knew the stock is going into the index then you are probably too late to profit from it as someone else will speculate the stock is going to go into the index the week before and already move the stock.
Funds can't really avoid this, and to be honest they don't really care to avoid it. It doesn't affect them at all, they are just supposed to mimic the index. Though you do start to get into a strange feedback loop whereby the index fund that is supposed to track the index starts to dictate how the index moves, we'll call this the "index inception" effect:)
I'm sure you are aware of this and decided not to go into the details, but for the benefit of others it's even a bit more convoluted than that. The index does not "buy" anything, there's a company that manages the index and just decides on the weights of each stock. There are then several mutual funds and ETFs that track that index who will actually have to buy/sell stock. So 3) is really
3a) how much the index weights will change
3b) which funds are tracking the index, and what's their total market cap
3c) what are the net outflows into each stock resulting from previous 2.
And sometimes a stock might be entering an index but leaving another (e.g., moving from Russell 2000 to Russell 1000), so the net result might be opposite to what you'd naively expect.
So yes, it is possible to make money doing this, but it's not a sure thing.
The very act of noticing that something is a good strategy, and beginning to trade on it, will over time drain away the utility of the strategy, until it is useless or worse than useless.
Tracking indexes is "big", and has some brute simplicity about it, but eventually the market will eliminate that as a viable investment mechanism. (But that won't stop your metaphorical Dad from swearing up and down you need to buy index funds....)
> But that won't stop your metaphorical Dad from swearing up and down you need to buy index funds....
The argument isn't that passive index investing is some kind of perfectly optimal investing strategy, it's that it's the most practical strategy for 98% of normal small, individual investors. If you don't have millions of dollars to invest, and you have a real job that prevents you from spending all your time researching investment opportunities, then you are probably better off just buying the market, instead of flailing around paying fees and trade commissions trying to beat the market.
If you believe in the weak form of the Efficient Market Hypothesis, and you are not a professional investor, then you should probably be in index funds.
Fees are the other key factor. Most investors are really using 401k accounts with captive choices, not a real broker with access to funds and free ETFs.
You're almost always better off with an index than the "HR Director Got a Kickback Growth Fund"
You seem to be misunderstanding the implication of anti-inductivity and/or the point of index funds. Anti-inductivity only holds that strategies that return better than average can't work when applied broadly. There is nothing that implies that a broadly-applied strategy can't give you the market average. Index funds are supposed to give you the average return, that's their objective. There is no reason to believe they won't continue to be successful in doing that (if everybody just invested in index funds, everybody would make the average return...)
Bogle started the First Index Investment Trust on December 31, 1975. Bogle founded The Vanguard Group in 1974; it is now the largest mutual fund company in the United States as of 2009.
The strategy is so passive that it will simply always follow the index. So in order to break it, the indexes themselves would have to break. I suppose it's not impossible, but it is very improbable.
Show me an actively managed portfolio that consistently beats an index fund and I'll believe it. Until then, you guys can pretend to have all the inside information you want, but numbers don't lie.
There are many private portfolios that "beat the market." Of course, you can extend the timeline or find myriad other ways to exclude them. It's very hard to be a hedge fund and beat the market because they do not have the luxury of sitting out poor market conditions.
I'm not suggesting "anybody can do it" but the study that concludes money managers underperform the market often gets stretched into "nothing beats index funds".
Yes, the top funds consistently outperform the market. No, the top funds are not the same year-to-year. Predicting which funds will out-perform is the hard (aka impossible) part.
Don't be fooled by randomness and survivorship bias. Those aren't the same funds each year. Thousands of funds trying to beat the market at a seemingly-random game? A small few will get lucky.
I would be more interested in whether it will outperform in the future, and whether it will outperform before I start moving into more conservative investments. Given that I'm going to be around a while, consistent but middling performance will eventually outperform inconsistent great performance. I would hate to invest in a managed fund only to have it underperform for several years.
I actually found $20 on the ground yesterday as I left the Bart. It just sat there, contradicting fundamental economic axioms like it wasn't no thang. And yeah, sure, I picked it up, but I'm also not quitting my day job.
Much riskier proposition because you have no idea of the timeframe of WHEN their positions will be divested. Your margin might get called before it happens and you lose or lose alot. With the long position buying the stock prior to its entry to the index you have a much more reliable indicator that you'll make profit.
Honestly the simplest thing to do is just monitor the big houses and whatever they do en masse do the same. Most giant purchases have to be announced and 10M shares aren't just bought on a whim.
This is well known known effect. S&P rebalances it index through out the year and there are always people trying to predict adds and drops and make some money.
The real diseconomy happens in the Russel indexes. They are rebalanced annually with the methodology for adds/drops announced ahead of time. Various funds that are pegged to Russel are forced to rebalance at this time buying and selling huge baskets in one day. To avoid large stock market movements and capitalize on them traders try to predict the changes to the index and prebuy the rebalance trade. Their actions through the market leading up to the rebalance and agreements to sell the rebalance trade to the Russel pegged funds reduce price swings on the day of the rebalance.
Trading desks that engage in the Russel trade spend the entire year preparing for it, modeling the methodology, acquiring clients for the rebalance trade, and prebuying the trade. Their profit comes from the difference between the closing price (mostly governed by Russel adds/drops) on the day of the trade and the price that they prebought at. Essentially their ability to accurately predict the rebalance add/drops and acquire clients to sell the rebalance trade to. There are desks that make $10s of millions this way on that day. There may be desks that make $100 of millions this way.
PS. I may not have stated it clearly, but funds that are pegged to Russel indexes make agreements with external traders to handle their rebalance trade for them at a fixed bps to the closing price. Traders are able to make money on this because they can take on risk and prebuy the trade; something that the Russel indexed funds can not do.
Even in the worst case -- Vanguard doesn't lose anywhere near this premium -- we're talking 20-30 bps, or .25%. It's one of those scenarios where one has to choose what is less bad. Sure, an active manager could play with the index a bit more to help avoid this, but you'd be paying a lot more than .25% for his effort.
It might, however, be a good enough reason to side-step this issue and use Total Stock Market (VTSMX) instead of one based on an index that frequently drops/adds stocks.
It's insane to buy one of these old index funds that are now pop curiousities. The only reason DJIA and S&P are interesting is because decades ago we didn't have computers, so we used extremely simplified indicators. Now you can buy into an index fund with 1000s of securities that doesn't have arbitrary constraints determined by the needs of historical contiuity with obsolete newspapers.
VTSAX (the Admiral version of VTSMX) is the gold standard in the early retirement community, as far as I can tell. Some of the lowest costs an individual can get, with widest US equity exposure possible.
But don't those Total Stock Market indices have some kind of weighting for each stock? Wouldn't changes in the weightings just leave you with the same problem? Or maybe I'm missing something?
They do, but Vanguard (and probably other fund companies) take a number of steps to help prevent frontrunning. For one, Vanguard recently moved its benchmark provider to FTSE Group (CRSP) for many funds, which takes more of an averaged trade-price approach to determining weights. They also delay trades and "packet" securities between indexes and buffering small changes in market cap from predicting trades.
It's not impossible for someone to squeak some money out of the predictability of an index, but it's not something that makes a ton of material difference with a company like Vanguard.
So, the writer of this article failed to give any proof other than citing the case of one specific stock (American Airlines). Even then, it failed to give any useful comparison (sure, the stock gained 11% in 4 days, but how did the rest of the market do?)
It would take a bit of effort to grab the data for all stocks entering the S&P for the last (say) 5 years, and then compare how these stocks did between the announcement and the joining of the index, but this data is vital to making the case that there is some market inefficiency here. Since the author doesn't bother to do this work, how can they justify their conclusions?
They can't even manage to get the 'easy route' right (letting someone else do the work). They cite 'one estimate' of a $4.3 billion cost but don't bother to tell use who made that estimate, giving the readers no chance to check its validity.
Lazy, lazy journalism IMO. At least quote your source, Bloomberg!
The source for the $4.3B is given later in the article:
"Over a course of a year, front-running -- of stocks going into and coming out of indexes -- costs investors in S&P 500 tracker funds at least 0.2 percentage points, according to research published last year by Winton Capital Management Ltd., a quantitative hedge fund that analyzed data from 1990 to 2011. That’s equal to $4.3 billion in lost income in 2014."
Thanks for linking that. What surprises me is that in their simulation which they base the results on - the early 10 years are way different than the later 10 years. The early years show fairly steady growth (although most of it is concentrated in the middle), and the later years (except for one outlier) are really flat.
This suggests to me that while it was a practical strategy, it isn't so much anymore. With index funds smearing their buys over long enough periods the effect shouldn't even be noticeable.
Is there a "Index Frontrunning" fund which I can invest in? (The fund would automatically frontrun all index changes, and so keep management fees to a minimum.)
I suspect this is more complex than just frontrunning. An article from the Fed Reserve Bank of NY from 2011 discusses the whole process of being listed on an index. Firms selected have been outperforming the market for several years at least, and their metrics have been improving even before selection is announced. This may be why these companies were chosen to replace others that are failing.
The Fed report found that there is no long term impact of being chosen to be on an index.
Sure some people get rich, but for buy and hold investors it's mostly a non-issue. The expense ratio of an S&P 500 index is still very small. A total market fund won't have the same front running issues and the expense ratios on those can actually be higher than the S&P 500 index funds.
The point of the article is that owning an index fund is NOT buy-and-hold -- the indexs sell to rebalance, and do so in poorly timed ways (that is, in hige fixed batches, contrary to standard advise to "drip"), exposing investors to trading waste that the indexes are designed to avoid -- undermining the purpose of the index fund.
0.2% waste is huge compared to the overhead fee of an index fund. VTSMX fee is 0.17%.
I did mean higher than, but that is because I incorrectly included the overhead of front running in the expense ratio when I think it would actually manifest as a failure to track the index. Some total market funds end up being a hair pricier (VTSAX and VFIAX are both .05% right now).
Still, if you look at VFIAX it tracks the index perfectly despite front running so it is still nothing to worry about. The amount of money in index funds is large so there is room for a few people to make some money without a big impact. There is some deviation that is significant around the 35 year mark, but do we even know that front running is the cause?
My understanding is that in practical terms weighting shouldn't matter for indexes for the most part since you should only have to buy/sell when funds enter/leave the index (that is the bug). The weight should track without any active trading because it's based on market cap. You buy it and if the market cap increases so does the holding and if it decreases so does the holding. No need for trading.
Now if you want equal weighting among the 500. Well that is an issue. It's one reason not to buy an equal weighted fund.
I know people are thinking of other more problematic indexes than the S&P 500. I don't because I don't buy them so I haven't given a lot of thought to what front running means to them. My investment objective is to hold the entire investable space weighted by market cap (modulo currency risk and home bias) and most indexes play no part of that.
>>I think it would actually manifest as a failure to track the index
No, because the stock was added to the index on the same day. So the fund is tracking the index, and the index is doing what it said it would do. It's just that they're both buying in an inefficient way that costs more than it might, and others capitalize on that inefficiency.
Right, but that's the fund manager's problem, not mine. A savvy fund manager would run a front running fund as part of its portfolio.
Also note this is only a problem when the market is mid-term upward or flat. The chances of this type of front running backfiring are pretty high in a bear market. Also, like all forms of arbitrage, there is an upper bound on the number of shares you can do this with before you're dumping more shares on the market than the index funds need. Granted that's probably a big number, but it's not infinite.
You do see PR from Dimensional from time to time. Their slightly smarter index funds probably do beat traditional index funds slightly but probably not after their additional fees.
How is this considered unethical? If you know X will buy loads of shares of an equity and you learned that fact legally through (presumably) public channels, are you A) a scoundrel, or B) smart, for buying shares before X buys those shares?
When somebody does something out of the ordinary it's necessarily considered a "game" (read "scam") by the majority establishment.
The reality is that the index is buying the stock because it views it as a good value, meaning that it is currently "under-priced". The index could be then viewed as unethical for "stealing" the shares away from current holders at a lower price than their intrinsic value, unless they were to tell somebody first... like a "frontrunner". In most cases, the largest entity is viewed as corrupt and evil, because it has the means to make the most efficient decisions and capture all the value, but somehow indexes are regarded as altruistic cooperatives.
They're not. S&P is owned by McGraw Hill Financial ($5 billion revenue), CME Group ($3 billion revenue), and News Corp ($33 billion revenue).
It's a 1%/99% story. The second paragraph pretty much says so: "hedge funds and Wall Street trading desks are reaping hundreds of millions at the expense of index mutual funds, the investments of choice for a growing number of ordinary Americans."
Wall Street vs The Ordinary American™ makes for popular news stories.
The reality is that the index is buying the stock because it views it as a good value
That's not how all indexes work. The FTSE100, for example, is simply the 100 largest companies on the LSE by market cap[0].
And many indexes (DFA being an exception that I'm aware of[1]) weight companies by their market cap, so the rebalancing trades made by index funds aren't based on value so much as an algorithm.
Just reading the title, I thought this would be another article about the Congressional loophole - where members of Congress can trade on insider information with impunity.
The reason this happens is because of strict ruling on how index funds must track the index (as the article mentions). There is no way around this unless...
You loosen the requirements on how strict the index fund must track. Portfolio managers are incentived to trade smartly if they are ALLOWED to. This is why bulk trades are usually "random" to avoid front-running.
So a solution here is create a competitor to Vanguard who has loose, yet well defined rules that define an index fund and how closely it must track an index. That's it.
EDIT: Others are commenting on index funds have to redefine themselves and add/subtract funds. But that is only the index, not the cause of the price movements. That is looking at it backwards. You need to look towards those moving the prices, like Vanguard, to find a solution.
Far as I understand it, the investment market exists on the edge between actively managed and passively managed funds.
Actively managed funds exploit inefficiencies in the market, but the more active funds there are, the less their returns through competition. Passive funds are wonderful when the market is efficient, but the more passive funds there are, the more inefficient the market, and the more profit active funds can make. This should create a natural equilibrium between active funds and passive funds where the market should "settle" and returns are optimal for both parties.
Mom-and-pop investors cannot take advantage of this (and many other opportunities) because they have to feed a long chain of middlemen through prohibitively high costs of trading as well as always being on the worst side of bid-ask spreads.
And then of course the value and the cost of high quality information delivered to you in a timely manner is a bit different than staring on CNBC screens.
Why can't mom-and-pop investors open an account at a place like Interactive Brokers and trade options on these events? The cost is pretty small per trade (maybe $1-2 for most stock and option trades) and the bid/ask spread isn't that bad if you are talking about an 11% move in a very liquid stock.
Just because if market does not move fast enough in the direction of their option trade - the value of their option investment will very quickly becomes ... $0.00
Because mom and pop investors have lives, children, jobs, bills, and the other stuff that demands their attention away from watching the stock market ever so closely.
When i used to very actively trade stock (even just thousands of dollars) i used to monitor releases from companies like MSFT's and Buffet's holding companies announcing upcoming stock purchases. I would then immediately buy the stock before they can. Then i would sell it a few days or few weeks after their purchase to lock in my 10-35% gain.
A way to avoid this is to not buy the indexes themselves but deeper capitalization based funds like VTI etc. This is still basically an index fund but since it buys past the "500" it avoids this BS.
Why would you want to avoid it? Either VTI is a better investment (for your profile, of course) than the S&P 500 and such or it isn't; it doesn't really matter if it's because of "index frontrunning" or any other factor.
Can't the fund "managers" just add is some slop in their buying so they don't cause such a large spike? The spike in demand is what makes this strategy possible, so spreading the demand over period of days or even a week should reduce the spike and the profitability of the strategy. I know they want their fund to match the index, but if this ever became a problem I think they could squash it pretty simply.
Is there a semi-passive fund that just responds to the announcements that something is going into or out of an index?
Presumably they'd leave a little money on the table because they'd have to guess how much to buy in or sell out, but it'd probably mitigate a good chunk of the loss, right?
It's much easier to do this with Commodity ETFs and Futures.
Doing it with equities or baskets of equities gets very complicated and if you read the prospective of most ETFs pretty much all of them track the indexes within some margin of error.
Commodity ETFs on the other hand have an underlying asset that will expire or require taking delivery. Most of these ETFs are managed by a group of less than 10 people. They simply do not have the resources to take delivery of an underlying asset. They don't deal with the hassle of storing / taking delivery unless things are really out of whack. Therefore, as a the underlying futures contracts approach maturity, they need to rebalance their portfolio, almost daily, and at minimum once every couple months. In a contango market, it is very easy to front run these funds compared to ETFs that track equity indexes.
I should know, this was my primary trade 4 years ago before quantitative easing killed all volatility in the market.
That being said, the Russell 2000 index is rebalanced only once per year. Opportunities to front run that index with the futures contracts are one of those trades that I miss dearly.
I personally know 3 people who run their own money ( <5 million ) who do this as their sole form of income.
Having said that, this isn't exactly easy. You need to know
1) if a stock is going into the index
2) when its going into the index
3) how much the index will buy
4) how much the index buy will affect the price of the stock
the first 3 are trivial for some index funds, though most have rules that allow them some leeway here so there aren't always sure things.
The 4th is where you make your money.
And it isn't like there aren't other's doing this, the article makes it seem much easier than it actually is.
If you play the game theory through you'd realize that if you knew the stock is going into the index then you are probably too late to profit from it as someone else will speculate the stock is going to go into the index the week before and already move the stock.
Funds can't really avoid this, and to be honest they don't really care to avoid it. It doesn't affect them at all, they are just supposed to mimic the index. Though you do start to get into a strange feedback loop whereby the index fund that is supposed to track the index starts to dictate how the index moves, we'll call this the "index inception" effect:)