7 ms·
One of Wall Street’s Most Popular Trading Strategies Is Now Failing
- zelly 8y agoToo many people doing the same thing in a zero sum game—of course it wasn’t going to work forever.
- 3dinfluence 8y agoThese strategies have only been around while the fed has been heavily involved in propping up the US economy with artificially low interest rates and quantitative easing. So I am not surprised that the algorithms and models that have grown up during this period are now struggling now that the fed is trying to back off a bit.
- FabHK 8y agoI don't think that's right. CTAs have been around since the 1970s. Momentum trading has been around basically forever. The Fed under Paul Volcker raised rates over 20% in the early 80s to head off inflation. Before the crisis of 2008 rates were 5% or so. It's only after the GFC that rates have been lowered to near zero. Also, according to the article momentum trading failed to deliver after 2008, while rates were low (and the market tripled), not just now.
- jhoechtl 8y agoThe interest rate based system of growth?
- djakjxnanjak 8y agoGood trading strategies always stop working (after fees) when they get commodified.
- mamon 8y agoReally? I always thought the opposite is true: trading strategies work BECAUSE they get commodified. I mean, it's like self-fulfilling prophecy: if enough mutual funds managers look at the price chart, draw the same magic lines, apply the same magic formulas, then they reach the same conclusion "buy this stock". And then they start throwing their billions USD of assets at it, and the stock price actually starts climbing up.
- adrianN 8y agoBut people who have better insight into the real performance of those companies can make more money if more people just apply magic formulas.
- askmike 8y agoThe most successful fund for the last decade(s) fits more one the magic formulas side (the Medallion Fund from Renaissance Technologies). EDIT: Oops, I misread your comment.
- simonh 8y agoI think both can be true, depending on the details of the strategy and the conditions in which it’s valid. Sometimes common knowledge of a strategy can lead to the elimination of the situation in which it works. Participants can anticipate it and so the conditions no longer occur. In other cases so many participants pike into the strategy, in which there is only so much money to be made, that returns get so thin that only the very most efficient and smart participants can get anything out if it. Arbitrage is a good example, once its known it can get traded out of the market very rapidly.
- allahreturns 8y agoThis is precisely what's happening with dropshpping. Now soon many people are interested in that, making money is lately just too difficult and most are making money selling courses to gullible people.
- quantummkv 8y ago
- Nasrudith 8y agoEspecially if they are effectively arbitrage - if it is easy money to buy 50 cent widgets, pay 50 cent for transportation and sell them for $5 in an area where they are for sale for $10 then someone can make a further profit by arbitraging your new price - and that is before supply and demand come into play.
- askmike 8y agoAt the end of the day arbitrage is zero sum, there is a fixed amount of money to be made from this market inefficiency (based on your margin, your costs and the amount you can buy and sell). Regardless if you are the only one doing it or not (if you are not you are competing for a piece of the pie). Note that the strategies described in this article are not arbitrage nor zero sum.
- djakjxnanjak 8y agoConsider the market for a bond. Smart investor figured out that the default risk on the bond is lower then we thought. More people find out and there's a lot of demand for the bond so the price goes up. Now the bond trades above face value, so new investors get a lower % return, and the bond is no longer a good deal, so the price stops rising. The same issue holds true for stocks, which are also valued based on a claim for some predicted future cash flows.
- pertymcpert 8y agoYes, and for anyone interested your second paragraph is the definition of a “security”.
- csomar 8y agoI thought that is the definition of an "asset" which is a broader term.
- PakG1 8y agoAssets cannot by themselves generate future cash flows. If a house just sits there, it cannot generate cash flows, it needs a tenant to do that. A hundred dollar bill cannot generate cash flows by sitting under your pillow. It needs to be put to work to generate cash flows. Assets can be used to generate cash flows, but they cannot generate cash flows just by existing. Hence the difference between asset and security, though both can be ascribed to have value.
- evrydayhustling 8y agoSecuritization is about ease of trading - which is helped by but not defined by ease of collecting cash flows. An asset is exactly something that can generate future cashflows, even if only by being sold. A security is a special kind of asset that can be easily traded on paper without changing it's value or taking on transformative effort. So, your goose that lays golden eggs is indeed an asset, but it might be hard to trade because many folks aren't well prepared to own a giose. If you wrap it up in contracts about someone who will house and feed the goose, collect and sell the eggs and forward the money to the holder of a deed - congrats, that deed is now a security.
- galaxyLogic 8y agoSo the article seems to be saying that CTAs are failing which is true based on numbers and then claims this is because the environment is too random. But shouldn't we consider as possible cause the fact that CTAs became common. When multiple parties use the same strategy it can not be winning any more for everybody.
- evrydayhustling 8y agoYes - and there is also a second order effect where latecomers to a well known strategy must focus on marketing and cost reduction to compensate for lower alpha, so each dollar flowing into the strategy is more likely to go to a less innovative, follow+on implementation (such as algorithms packages offered by big banks).
- plouffy 8y agoThe article notes this: > Part of the problem, according to critics, is too much money now chases the same trends, undermining what traders call "alpha" — the outperformance possible relative to a benchmark.
- pfortuny 8y agoPredicting the future works until is does not. This happens a lot more frequently than ome might think.
- chvid 8y agoMarket moves sideways - trend following underperforms ...
- ashelmire 8y agoFrom the graph in the article article, trend following has been abysmal for the last decade. > Between 2008 and 2018, Societe Generale’s main trend-following index made only 3.7 percent, compared with an average gain of 62 percent for hedge funds and a more than tripling in the S&P 500, including dividends. The market performed incredibly well over this period. These algorithms didn’t even keep up with inflation.
- skookumchuck 8y agoAny algorithmic strategy will cease to work once it becomes generally known.
- ackbar03 8y agoI think getting a bunch of smart people to work on making automated trading strategies 24/7 is a tremendous waste of talent
- 2sk21 8y agoThat and figuring out which ad to show you...
- wizzairflyer 8y agoI see your points but then the question becomes how to give them an incentive to leave those fields. It feels that they are the only one that pay and consider you to your true value.
- Joe-Z 8y agoNo, that does not become the question. They chose money over morals, that's it. It's not like they are helpless victims to economic principles, so they have no other choice than to work for the scummiest types of companies because they offer good pay.
- samatman 8y agoSure, you can stop there, if your interest is in moral judgement. If you’re more interested in shifting the landscape, changing the incentives becomes the priority. Only one of these puts a stop to the objectionable behavior and wasted talent. So I know which I prefer.
- cousin_it 8y ago> If you’re more interested in shifting the landscape, changing the incentives becomes the priority. Or changing laws to make the bad stuff illegal. I'd be happy to make most ads illegal. In some places they banned all roadside billboards and you know what happened? It made the view better and didn't hurt the economy.
- speedplane 8y agoIf an algorithm existed that could predict the future value of stock prices, that same algorithm could be relatively easily adapted to see the future. Not only is that tech impossible, it's also depressing to think that a true Oracle's primary function would be to help traders make more money.
- frozenport 8y agoBut does the future it sees include the purchases its about to make?
- speedplane 8y ago> But does the future it sees include the purchases its about to make? Yes, absolutely. Go back to Greek mythology (or even the Matrix movies). Many predictions the Oracle made only came about precisely because the Oracle made those predictions.
- piker 8y agoWhy does this article call them CTAs? Being a "commodity trading adviser" is a CFTC regulatory status, not a trading strategy. Those issues are orthogonal. One would have expected them to be referred to as algorithmic traders or algos or something.
- rorygreig 8y agoMy understanding is that CTA is a general term for a hedge fund that uses trend following strategies, primarily trading futures, but not actually restricted to commodities, so it's not a particularly accurate name, just industry jargon.
- rorygreig 8y agoThe fact that trend following strategies are becoming less effective could be interpreted as a positive thing implying that markets are becoming more efficient, as trend following shouldn't work in a completely efficient market. Trend following strategies are pure market followers and are not based on fundamentals, therefore they do not bring any new information into markets and are unlikely to improve the efficiency of a market.
- NicoJuicy 8y agoI always thought that HFT is a losers game. There can only 1 be the fastest and those that invest their money this way, could probably used it in a better way. And if you are winning. A small error in the algorithm could make you lose all of your profits. Rare events do happen and you haven't covered every case.
- zxcvbn4038 8y agoI once watched a bunch of card counting videos on YouTube. It didn’t matter who made them, all of them were basically N minutes of the guy cursing and complaining that the other players were getting the cards they should have gotten. This sounds like more or less the same thing.