"Combining the Best Stock Selection Factors" by Patrick O'Shaughnessy

you all right so we'll get started right about now so I'm very very happy to be able to introduce to you Patrick astronomy he is the principal and portfolio manager at O'Shaughnessy Asset Management he has also published a book called millennial money how young investors can build a fortune published by Palgrave Macmillan Wellington great uh-huh I got I got there he's also a contributing author to the fourth edition of what works on Wall Street so without further ado Patrick Johnson thank you so thanks thanks for joining me this afternoon so what I'm going to talk about today is stock selection factors in the equity world I think this is the important topic today because while this used to be a strategy which was very very fertile ground being if you had come across this strategy twenty or thirty years ago it would have been probably one of the best way to to outperform the market it has become very common ground arguably over farms or Overmind ground for trying to earn alpha in the US and international equity markets so I'm actually going to talk about the ways that the industry I believe has misused this idea of factor investing so broadly speaking buying stocks both long and short or building long short portfolios based on factors like valuation momentum quality low volatility being some of the most popular factors out there today and why instead future potential alpha in this style of investing will look very different it will look more concentrated so smaller portfolios of stocks that exhibit a lot of different characteristics and ideally stocks that have factors that have not yet been identified and commoditized away so this used to be a really really great way to put your money into a simple you know value portfolio and stop worrying about it and the question is is that sustainable for the future so we've spent the better part of 20 years or so running money so about five and a half billion dollars today on this idea of quantitative very systematic investing strategies and the idea is simple that you find key characteristics or common DNA of stocks that have tended to outperform and you just constantly rebalance towards portfolios of stocks that have those characteristics in any given market most of this is done at first using back tests so you get big historical datasets financial statements price data on companies going back anywhere between five and eight decades you find the best stuff you build portfolios around them having spent with a couple of the other presenters here today I'm sure more time than I probably should have with this data through history I can assure you that that testing looks a lot like this funny cartoon where you see great results at the top level but once you start actually digging in to what's driving those results oftentimes it's micro cap stocks that you can't trade as an institutional investor where there are data issues or huge sampling issues there are so many problems with building a strategy around just a back test and unfortunately that's become probably the most common way for big institutional asset managers to market new strategies because it's very low cost for them it's very easy to set up a rules-based system that does something with a portfolio manager at the head of it who can somehow manage portfolios of 10 or 15 or 20 or 30 different ETF's so it's a very scalable business model to build you know an alternative to a simple market cap weighted index but unfortunately I think that because everyone knows about these things the excess return that the strategies have earned in the past may be to some extent jeopardized in the future I certainly don't think it will go away and I'll describe why but I'll walk you through kind of the evolution of this idea of factor investing and how we think you should instead use factors to combine into an overall portfolio so first of all this is the kind of level set with this recent trend or deterioration I'm talking about with factors this is five very big broad widely accepted factors that you can get this data's on fama and French is website sort of the big five if you will for investing strategies for factor investors clamed some operating profit which is basically a measure of quality investment which is kind of a counterintuitive one you actually want to buy firms that have following long term asset basis and you want to avoid firms that have the biggest ramp ups in their ass basis so you know biggest growth in goodwill by groping Catholics you know instead of a tend to be very expensive size small caps a little bit ascended outperform large cap and value book the price being the most common factor there and what you're seeing here these lines going all the way back in many many decades is the annualized rolling ten-year long short spread between best and works so you can see for this kind of whole initial period let's page both the price has the example that's that's the purple line you're earning this kind of incredible and very steady long short spread by doing nothing more to being long the cheapest docs cheapest F model stocks my price the book and short the most expensive desk I'll sort of an easy way to earn incredible longshore returns and then you can see what happens that kind of broken out by interest rate regime so rising rising secular falling secular and then this zero interest rate period but more importantly what you see is this massive deterioration of these long short spreads so pretty much all at once these factors that were sort of the one markets version of a free lunch if you could have somehow identified them many decades ago have stopped working to a considerable degree versus their long-term history and this should be very alarmed because when someone knows about something when everyone is agreeing on one way of investing money oftentimes that gets rid just happens with stocks oftentimes they get very expensive and any edge that you may have had is very quickly eroded in the overall market so when you see this kind of erosion you should you should take pause you should wonder whether a strategy like this is going to perform in the future there are very specific reasons why some of these factors have failed so I'm not a believer in the size factors you can see here that is certainly the least consistent the blue line both Mullins actor historically and we happen to be sitting at the tail under the time right now you know in 2016 we're large-cap since oh nine has crushed small cap so it's been a great time to be a SP 500 or an index investor but the rest of these are pretty solid factors especially these ideas of value momentum and quality I would define quality very differently than plum and French and kind of the academic literature does but those are really good well proven long term ideas for investing by Chief stuff that's not completely junk quality you know not over lever to the hills and buy it when it's trending so some can stay cheap for years and years wait till it has a bit of an uptrend just the beginning legs of a positive trend of the market and buy it that's been a really really solid strategy but they can suffer and you can see this deterioration happens fast so this is the average longshore express of 10 percent on average really good at rising and falling rates but more recently very small there's an amazing book that if you do anything from this talk what you should do is go read this book it's called a misbehavior of markets by a guy named a fractal geometry mathematician named Benoit Mandelbrot it is the best book you will read about markets and it's one of my favorite quotes is that the trend isn't has vanished killed by its discovery that's what happens in markets all the time when there's a big advantage people become aware of it they put a lot of money behind it that bids up the valuations typically of that idea or factor and then it doesn't work in the future mean reversion is a as they always say book to price I think is the most interesting example of the law that I that's called good hearts law so good arts law says that when a measure of something becomes a target it ceases to be a good measure so there's a lot of examples of this maybe the most tragic one in the u.s. is that homeownership for decades and decades was a measure of happiness a quality of life it became a target that Congress and and the financial world sought to achieve for everyone and you know how that ended my favorite examples were in in colonial India and Vietnam so in Vietnam they had a huge rat infestation problem and so the government said we're going to offer a bounty on rats if you deliver a rat's tail meaning you've killed it we'll give you 20 bucks so what people did wasn't kill rats they just chopped their tails off turned in the bounty and let the rats go back to breed so that their bounty pool increase same thing happened in colonial India it was with Cobras they said bring up a Cobra and we'll give you this bounty well what people did was they started breeding cobras cobra farm they were not they would breed cobras kill him turn him in for the bounty and then the bounty payers realize this was a terrible plan said okay we're not paying more bounties they were the breeders released all their cobras and the population was bigger than when they started so there's always these unintended consequences when a measure becomes a target and arguably price the book that priced the book factor which was originally held out by fama and French in 1992 as the seminal value factor it had a lot of great features at the time and you had looked at a menu of value factors in 1993 which I'm sure if I'm in French did you would have seen that sense of the early 60s value factors had all provided a pretty similar amount of historical annualized excess return so you can choose well earning sales even da whatever you wanted to choose it was a pretty effective strategy no matter what they chose price-to-book because it's got the lowest turnover so you could build a portfolio around it turns over 1020 percent a year lower trading costs better tax efficiency and so on and here's what's happened to those same factors of the Soudan excess returns sense common french published their paper so exact very similar clustering of returns prior to the paper hundreds of billions of dollars including a couple massive asset managers that I'm sure everyone in the room is familiar with and build strategies value strategies based on this idea price the book and you can see that since that time it's been by far the worst performing book factor or value factor you hear all the time as we do on my quarterly calls with clients that value has been kids getting killed by growth and the reason people say that is because of price to book they talk about the Russell 1000 or 2000 growth versus value where growth has won but if you would use anything except for book for that same calculation over the last ten years which is this period people are talking about that story goes away if you had SWAT sales in for book for example for the last ten years it returns would have been a hundred percent cumulative least better so it's always a concern when a strategy when a factor when an idea when anything becomes too popular and becomes a common target for asset managers because very often what you see coming out of that popularity is an erosion of an otherwise strong historical advantage the problem with momentum so that value the problem with momentum is that it can be absolutely excruciating to live through certain periods certainly long short but even as a long only momentum investor these are inversions in momentum that all has a similar feature back to the 20s all these circle periods and this is not an exaggeration so the long short spread with negative spot levels negative 500 percent in one year for momentum coming out of the Great Depression in 2009 and I brought nomensa portfolio so I can assure you this was excruciating long short momentum underperformed by about 200% so what happens is the market goes down 40 50 % route market realizes that the world's not going to end the stocks that were priced like the world is going to end bounce off the bottom think forward Bank of America City etcetera and 2009 they're up four or five six seven hundred percent your short those stocks you're along the stuff that had protected into the decline which were barely flat in 2009 so you have these unbelievable inversions in momentum if you think back to that first slide that you saw momentum tailing off as an effective factor that's entirely because of really like a five month period of time since then momentum is actually held up very well arguably it's been the best performing stock selection factor but momentum can be exceptionally painful this is true of most of these factors in general it takes huge discipline to profit from these things I would argue that most of the profit from factor investing has gone to sponsors or fund managers who use these strategies because they're selling raising assets in strategies charging fees and doing very well for themselves the investor returns if you will time weighted versus dollar-weighted returns have been horrendous so even with clients of ours quite honestly now we've worked hard to build a much longer-term capital based clients that really understand what we're doing but like any strategy people pile into something after it's done really well a pile out after it's done badly that has been the big issue with factor investing like any other style has been has been that people have not been able to stick with it because you have to live through stuff like this so I'm going to talk to you instead about a very different factor because I think the goal of anyone that's trying to do systemic investing especially in equity markets is trying to do something a little bit different and ideally do something that is not scalable that cannot cannot be commoditized the way because there are not enough companies that meet the criteria of type that I'll talk about so value momentum very popular investing strategies one that we are well known for and kind of hang our hat on is this idea of capital allocation so this is really getting it's going to sound coming to sound up maybe night I'm at the wrong conference not a Quan conference but a fundamental conference it's going to sound like a fundamental analyst might look at a business but we like it from the perspective of a quant because when you think about capital allocation there's a fairly limited menu of options so we can test how companies mix these options sourcing and spending capital so how can you source capital well you can get it through normal business operations that's arguably the best kind internal financing we call it you can use that you can use equity and you can use float so maybe floats the best its interest refinancing payables or buffets traders for using flow to fund Berkshire Hathaway that's really a on the spending side you can spend on capex Rd you can acquire the businesses paid on debt pay dividends repurchase shares that's kind of a major menu and the question we have asked ourselves is what mix of those different options has worked out best for the shareholders of those businesses so values great Momentum's great but what about how businesses allocate their capital because they're investors too just like we are this is some interesting very very middle e very high level results for some of these categories these are annualized excess rates of return good and bad versus us large stocks that this isn't a small is no small cap effect in here the big US company companies that dilute their existing stake holders issue a lot of equity issue a lot of new debt have underperformed between two and four percent per years companies is a big what we call Empire builders so biggest growth in their long term asset basis both capex biggest aggressive expansion and often as a long time at the peak of cycles and companies are the most acquisitive often again at the peak of business cycles have actually underperforming that one often surprises people the most because you think well investments for the future you need to spend often to grow and if you want long-term growth you need capex capex isn't evil but the extremes of capex growth do tend to be evil so you want to avoid companies with the highest rates of in percentage terms year-over-year growth then you get to finally some good excess returns which is businesses that we would argue have more discipline capital allocation or returning cash to to by paying down debt and they're spending a huge amount of money on share repurchases this next slide always gets botched I'll skip it the idea behind that slide was and probably the most important factor in all of this capital allocation is share buybacks which you probably read about in the news it's become a popular whipping boy of corporations buying back their shares and most often painted in the negative lights that they should be reinvesting that this is financial engineering they're issuing debt to just buy back shares it's almost always negative what we find is that for the most part we agree but there is a small sliver of stocks and again remember I said you want stuff that is not scalable you could not shares could not launch an ETF that could accommodate much money in this kind of strategy stocks that are repurchasing a massive amount of their shares outstanding net of any issuance none of any goofiness with employee options or anything like that that also created very cheap prices with decent quality has been an incredibly powerful combination for stock selection historically so companies that have a more disciplined capital allocation that aren't overspending and that are recognizing a current discount to their intrinsic value by repurchasing sometimes 10 15 percent of their shares outstanding that category very overlooked and has tended to do very very well this is a distribution of the relative prices kind of maybe too complicated for this but I'll go through it it's the distribution historically of the relative price paid by different companies buying back their own shares at different levels of conviction so let's say you're in purchasing a billion dollars of your own stock if you're Apple arguably that's a very low level of conviction in your stock it's one six hundreds or whatever the market cap is today of their overall market cap if you're Parker Hannifin or Northrop Grumman or a much smaller say 10 billion dollar stock repurchasing a billion dollars arguably you're making a much bigger bet in your own share price one tenth of your total market cap in the exam the ten billion dollar stock so we defined conviction between zero and five percent of shares five and ten and then ten plus ten plus being very rare but exceptionally high conviction this is a distribution of all our stock so it's flat this is a valuation anything plotting here would be the cheapest five percent of stocks in the current market most expensive five percent of stocks right now this these two lines are the high conviction group and this is the low conviction group so the low conviction has almost no skewness here towards preferring to buy back stock at cheap prices they're just kind of doing it because the most large stocks these days are doing it it doesn't seem to be evaluation play they do slightly avoid buying back expenses but look at the skew or the preference for high conviction buyback firms to be repurchasing their chairs at very cheap prices this is exactly the kind of categories of stocks that were asked her because what they're doing is saying let's reduce your account by five ten fifteen percent but do it while we're trading in the chiefest five percent these RPGs of six seven eight you know price to sales of point three point five really cheap relative prices versus the overall market and there has been a very big difference historically in the returns of these different kinds of businesses so all large stocks over the last thirty years or so eleven percent return net issuers so companies that are diluting their shareholders that underperform by about a percent the low conviction buyback firms have outperformed by about a percent but the high conviction category has done much better outperforming by about 5 percent per year annualized again versus large cap stocks is the hard market could be arguably if markets are efficient to any extent the most efficient market in the world is probably the US large cap market so an incredible degree of excess return for a pretty interesting factor that typically is not represented in a portfolio of someone that calls themself a factor investor this works exceptionally well with value so a quick word on value investing which is probably the most popular form of quantitative investing because it's been around the longest and it has great philosophical roots with guys like Ben Graham and Warren Buffett we think of value investing as as putting yourself out there as the house in Vegas the house in Vegas the big blackjack as an example has a one new a half-percent edge what they do is they turn that very small edge into a massive cumulative edge over time through volume right so they're know they're going to win over time because the odds are stacked in their favor in the short run they can get destroyed right so someone can go on a run and we do the Grouse stock as an analogy here of Facebook or something like that extracting or a gambler can go on an extreme run in the short time but over the long term the house is going to win value investing looks very much like that so actually only about 51 52 percent - just like blackjack percent of value stocks as we define them have beaten the market historically but if you let that in shorter periods of time if you let that compound over time we've actually never seen 10 or 15-year periods where these kinds of stocks have underperformed the market as a group so if you bought a big basket of them this is the distribution of value and growth the value is much more peaked this line is zero so anything to the right of here is our performance I mentioned that about 52% for the 8% growth as a PES the bad kind of distribution that you want to avoid but a fatter table effect so there's more examples of growth stocks doing exceptionally well arguably this is what keeps people coming back to the casino tables in Vegas it's the same thing that keeps people coming back to buying stocks like Facebook like Tesla today you know pick your favorite example almost the first question we get on any call or talking to advise or anything like that that might use a strategy of ours is well well this process systematically miss stocks like Facebook and Amazon and Google and so on and the answer is of course yes it will but we don't care because we're not interested in individual outcomes we're interested in putting probabilities on our side so by definition to do that you have to acknowledge that you're going to miss some unbelievable returns on the right-hand side of this distribution but you also know that you're stacking the odds in your favor so value works really really well interesting that when you look historically at the returns in the earlier talk net favor was saying that one of the great things about value is not just that cheap stocks outperform maybe even more so it's that expensive stocks underperformed by lunch so if you're willing to build a portfolio that's not just a tilt of the market but is a true often painful painful to buy and often painful to own the value portfolio not only you're getting longer term returns over here but you're avoiding the stocks and tend to fall in this category yeah question valuation metrics of living so it's a combination of for its sales earning EBIT and free cash flow just equally weighted so those factors like a full month yeah sure 5x is that it's moving like the last whole month share buybacks or coming to now share buyback plan you what they claim you're going yeah so same look-back period twelve months you can do it shorter you can do six months there are blackout periods for when firms can repurchase their shares you don't want to go to a month or something like that you can go back to further two years but one year really does work best because you want actual activity so you definitely do not want to act on announcements there's studies that show you know 26% of announced buybacks ever actually happened so it used to be a signaling tool as you see got a little harder on firm through purchasing shares so now they have to disclose it monthly and that 26% come up but you don't want to do it on announcements as a 12-month look back so that's B that's the distribution and so I wrote a book called millennial money that's focused on young investors and maybe one of the issues is that new generations keep making these same mistakes over and over again so what you're going to see on the next slide is the most owned stocks that have the youngest median owners and where they plot today along this continuum so these are the stocks that young people own the most done and where they sit and kind of the relative excess returns that those categories have earned historically Apple arguably Biscuits the flows so big they have enamel but all these stocks are extremely popular and fantastic companies but there is a huge and important distinction between a good company and good investment and price is definitely squarely on the wrong side of these kinds of stocks and then as people get older so these are the stocks and have the oldest median owners so it's not the most boring portfolio I couldn't I couldn't strip the more boring pork oil than this but much cheaper on average than stocks that young people tend to prefer so maybe this is just a big generational wheel wheel torture that people don't realize how later in life do they reach into whether they're going to be selling their stock that they should own cheap stocks is that the commentary on also just a way to market that the things of the same values asset because to me I look at this and say you know can you go back to the basics of Finance instead of ETFs and what what what it is whatever we put numbers on what the business should be work guide an aggregate is correct and that's always the thing because every time you pull were paying for some future future return and you don't get them and every time you take some sales value or you know for median you can to kids or weeks on you expect so there is an element of I think that that's adventure isn't it's an overused quote but it's right which is that in the short term it's of voting machine in the long term it's a weighing machine and the problem is that that weighing machine is very long-term so over the over a multi-decade period yes stock returns tend to follow basic fundamentals like earnings growth but for long periods of time they can have huge swings around earnings growth rates and so for all of you who are probably sure because you're interested in now performing markets or in systematic trading strategies I think things that take into account psychology price sort of current opinions that tend to overdo it at extremes is the far more interesting way of profiting from all of this but maybe you know maybe it may be just laziness at the end of life you just touch it less and it tends to get cheaper more seasoned companies so one interesting thing is that value which again using the Russell test proxy has underperformed is actually interesting starting to look a little bit interesting again so one of the things that we look at is the spread between cheap and expensive stocks so arguably you want that spread to be wide right so the value strategy works when spreads are wide and they start to narrow rights evaluation ratios converge for a lot of time in the last you know three four years it's been very very low my assets are close but low and then in the most recent period a lot of this is energy and materials but they have started to spike up pretty aggressively so may be good news for longer-term value investors that we might finally starting to be able to take some opportunity buy some cheap stocks again and expect to outperform here's what happens when you combine these ideas so tie shareholder yield we like cheap prices we like if you buy high Sheryl yield and and cheap you know good returns 13.7 percent you much worse if you buy high shareholder yield at expensive prices so it's a high dividend yield and a lot of buybacks but it's otherwise really expensive that's been a really bad sign historically you want to avoid it the way this all then rolls together into a strategy and I'll close talking about how you actually build a portfolio these are the things that we look for cheat nets let them talk a little bit about quality I won't go into but a whole other kind of more fundamental oriented looking at balance sheets looking at accounting choices kind of preferring high cash earnings etc and then this idea of shareholder yield so we build portfolios around that the difference is between us and most quants and probably the biggest philosophical disconnect is most quads famous and otherwise that you get in here would say that factor investing or quantitative investing is all about broad diversification and averages that you want big broad exposure to hundreds of stocks long and short so you're not taking any individual name bets you get higher information ratios doing it this way you can hedge out industry risk all this other stuff we actually disagree we think that the future potential for these kinds of strategies is to be active so there's a concept called active share which has become very popular in the in the equity world which is basically a number zero to 100 that measures how different you are from your benchmark so let's say it's the S&P 500 it's an active share of zero is news next month it means there's a 100 percent overlap between you and the market an active share of 90 means you are extremely unique means that there's only 10% overlap between your fund or hedge fund or whatever and the S&P 500 now the industry's trend has been towards lower active shares something called closet indexing so there's this huge problem in our world of principal agency risk I'm a fund manager I want to keep my clients and to keep charging fees to run a good business if I have a more active portfolio which opens up the kinds of cone of potential excess returns good and bad if I strike one of those bad years the odds of my being fired go up very quickly right and this agency risk has created this trend towards closet indexing where people build portfolios that look more like the overall market it probably means they're not worth the fees that they charge but it also means they're less likely to get fired in any given year active share is not a predictor of future excess return but it is a good predictor of potential future excess return so what this is showing you is you ran an optimizer to say okay in any given year this is just the average of all those observations at different levels of accuracy heirs of 10 20 all the way up to 90 what is the maximum possible excess return that I could earn in any one year do you see that there's always an upward sloping line and Ferries but the average is that as you get more active your potential gets better and better so if you're going to hire someone that's a 20 30 40 50 percent active share manager you should probably arguably pay less for that manager because their potential is is is less now this cuts both ways right so the same is true on the downside that more active the more under performance you can have versus the market and this is actual performance of all mutual funds over several decades by their active shares so you can see that Co now is talking about where it's got it you know if you're an index money obviously going to look just like the market as you get more and more active the chances of doing really good or really badly white and arguably if you're going to pay for active management which fewer and fewer people are doing you want at least the potential of earning excess return in the portfolio and then the question is well how do I get on the right side of this distribution curve and we think the answer is factors that are a little bit more unique different takes on quality and value this idea of high conviction buybacks etcetera find your own factors that's that's the whole magic of all this research world that that we've evolved into but this cone is really important and then that result is that we think you should go at different ways in the industry has so every one those morning star charts would show you a little dot that the average usually it's small mid large and then it's valued core growth and it's a little dot that says okay here's where plots on average and then the circle is 75% of the funds Holdings so smart beta you've probably all heard that term it's mostly the most these factor investing strategies or term smart beta they have hundreds of Holdings a lot of overlap with the market mole active shares the charge active fees or at least more active fees these are some smart beta options so we've got the Russell 1000 that's just the market Russell one value arguably the first smart data strategy tilts towards value you can price a book and then a very popular one the research affiliates called the rocky 1000 which is tilting again towards smaller and cheap stocks you see you know a little bit of a preference for what we want which is down here Chris you're older yield great value quality etc but a very broad exposure to the overall market we think you should build instead portfolios like this much more clustered around stocks with the best overall characteristics which could be as few as 50 holdings in the overall portfolio it gives you a much tighter exposure to the things that have worked well historically the downside and probably why you don't see more of this is that it's not a scalable strategy so this cannot accommodate the kind of assets that the major sponsors of ETF and even mutual funds want right so you want to launch something that has the potential to have 50 billion dollars in it to be able to charge fees on this is not that kind of strategy even in large cap it can still can accommodate billions but not not tens and tens of billions and its small cap you're talking maybe a billion or two worth of assets strategy like this and I'll close by showing you a sequence of plot points where this is where the industry is gone and then I'll show you what we prefer this is every stock in the S&P 500 where plots on a continuum so this would be best down here at least in our way of looking at the world and the size of the dot is the weight of that stock in the index so there's the S&P 500 there's the Russell 1000 value so you can see a little bit of a tilt start to happen you see left in the upper right so you had some bigger weights over there that the Russell 1000 value is avoiding but still very broad exposure there's fundamental index so again a little bit more of a tilt towards that bottom left there's a very popular ETF these days called minimum volatility so very low volatility stocks which also happen to be as expensive as they've ever been in recorded history and there's an approach that we would advocate which is one that is again much more focused in the kind of narrow corner you see some outliers right so you rebalance the thing infrequently annually or less often usually the whole interior year and a half two years something like that so for example this is a stock hold excel group and insurance company had great shareholder yield was buying back a lot of shares in May of last year it issued a bunch of shares in a cash and stock deal to acquire another business so a month ago was a month ago that dot wouldn't like this down here and it's drifted and gotten smaller because the rebalance process is the kind of quantitative systematic process slowly reduces weight to names like that so this is the approach that we think will work in the future it may not work as well as these strategies have worked in the past because everyone knows about it has access to the data that we use but you would argue that because of the pain that it sometimes takes to stick with a strategy like this because it can be painful for three even five years sometimes of underperformance and because it's not scalable for the asset management business which is a really important part of what drive stock returns overall that a strategy like this especially for individual investors but for smaller investors is a far better way of using factors overall this is the kind of portfolio you get so is a 55 stock portfolio it's going to share what our yield of 922 percent almost more than almost four times that five times that of the index at two point one year oldest talk about live acts of the indexes five acteal is actually negative so when you encounter all the share issuance that goes on at the same time or basically no buyback happening which is pretty interesting very cheap right 12 times earnings one-time sale six times Aveda a little more on free cash flow but huge discounts to the market again if value works because people get too pessimistic the lower the valuation ratio the more pessimistic the better your expected future returns that is only achievable in the portfolio that looks very different that it's very active well that's the S&P 500 that is not the direction the fund industry has gone so those kind of advantages have led to very interesting results so I won't even tell you what the names of these strategies are so I'm not trying to promote them and I don't think most people could even access them but these two strategies is a small cap and a large cap version of what we've been doing for 15 plus years on a cumulative excess term just cumulative returns overall since 2004 which is when the small one watch so two hundred and forty two hundred and sixty three percent we show them next to we call it smart beta so this is just a Russell large and small style indexes so the major story has been growth outperforming value so those are growth those are value when you put in a lens of what can be achieved through a much more differentiated portfolio all of a sudden that little story seems to go away it was very clustered like they're all delivering kind of different versions of the same thing and I'll close with an example that I've been using recently because I just think it's the best way of illustrating this idea between whether or not a strategy is about actually earning alpha or whether it's about gathering assets for the sponsor of that strategy so the Russell 1000 value Marga bleah one of those popular value strategies out there one ETF alone has twenty five billion dollars in it the iShares one we're just talking the Russell 1000 value is ExxonMobil which is a popular value stocks look a lot of large value managers you'll see ExxonMobil in their trading low multiples of earnings which the garbage way of evaluating an energy stock to begin with but still biggest stock in the S&P 500 it's there because of its price to book which is the methodology behind the Russell 1000 value it's only in the 50th percentile of all stocks in the Russell 1000 value by price the book so it is far the weight of stocks is far more determined by their size and the derivative of that is the assets which can be accommodated in strategies which charge management fees to invest in them then is its price the book itself so if you believe in factors any exposure to the market cap factor they lose your advantage but that is where the world has gone because there's an industry that sits on top of us so I would urge you read that book by Mandelbrot and if you're going to go into the equity world using factors do it in a way that is agnostic to the market don't be a slave to the S&P 500 use factors to build a more differentiated portfolio because I would argue that's the only way of having a sustainable edge into the future thank you you guys implemented this in a long short models well succeed what happens when you short the printed ones and emulsion yeah so yes and we failed and and the reason is so you know you look at something like this and you think god perfect long short tool you know so long this stuff short that's up to problems this stuff is exceptionally volatile and out of the strength and often very just from a logistic standpoint operationally very expensive or possible with borrow so when you're talking about shorts paper or back tests when it comes to shorts or is a dangerous game to play because when you go and actually try to implement short size of these factors in the real world they can beast you pretty quickly that's been our experience I think there are very good ways of doing this it probably is in large caps where you start with the universe of easy or cheap to borrow stocks and you do a lot of matching so I didn't talk about like portfolio construction but you know do you take industry bets how big of an industry better willing to take we want to take oil that's there's all sorts of risks that get baked into a portfolio a long short application of this idea I think needs to very smartly and wisely manage risk exposures so you probably want to be neutral to industry you probably want to be neutral to big risk variables because those things can read think about you know energy as a good example if you were just long short value last year you got absolutely destroyed right because you got two ways you were long you know miners and and energy and stuff like that and you were short Facebook and Amazon and and those kinds of businesses you know David Einhorn is a guy I respect you know as much as any anyone out there felt that pain a lot of big famous guys felt that pain if you're agnostic to industry and things like that you can get it both ways and that is not fun to live through and often client money will not stay with you through the times like that so we've tried it frankly we failed not to say we won't succeed in the future but there are or there are big hurdles to doing it right yeah have you had any fears it's all sectors for example it'll be a bad situation with giant have come to Odin take that find out the number of parking lots which is basically by accident I know anyone much so and you were the mall parking lot for balance which is resolved so having experiences with soft factors and how they grow is the create expired but yeah so the question is experience with soft factors so we think about this as like unstructured data or they're just really unique things that might be specific to an individual company like how helpful is a parking lot at Walmart or something like that so the answer is yes and no we've dealt a lot with things like sentiment and news data where you use key words and machine learning type stuff to try to find interesting factors we don't use it so we've looked we've yet to find something that is we think worthy of adding to our process I didn't say this but one of the big important things with these strategies is not changing them too much because the temptation is always to rotate into the hottest new thing there's going to be an extra hat then and will continue to be an acceleration of the amount of papers being written on factors and a new one comes out and it's better than the fact that you bought a year ago and say well you know screw this I'm going to bail not want to go into this one that kind of rotation is really dangerous because typically papers get written at the tail end of a mean reverting cycle so the factors been doing really really well and then something structural you know people read papers about it and some in structural changes or tendencies change and it does really badly so a really important caveat to all this is there's going to be stuff that comes out after you choose the strategy that's going to test better than what you've got and you need to be very careful about making changes so when I say that we haven't added it we have a super high bar for adding something into our process and nothing has yet met that bar yeah bits of it yeah I I think probably 75% of this is in various papers and posts and things like that we could our our mandate is that we believe in focus so we are all equities 100% and have not delved into the option market we're not a hedge fund right so it's not we don't have it just a wide-open mandate we're hired very specifically to do one thing so that is certainly an option in a way around it and this is on all stocks or the small caps in here and I don't know enough shows will be expense they'll be expensive or what's off saying because it'd be probably in the liquid right differences at all different vectors I mean variances in which the women return those are because an acquisition factor integration so yeah and comply back so the answer is way too long to give here when I said that that one kind of simple back test I showed was high level I meant that every one of those factors has a lot of nuance and each of these factors has different information horizons so like momentum sir isin is very short you know three six months values is very long five from them to ten years and these different factors have different horizons and that's part of the prop that's part of the management process so I'll stick around and we can talk a little bit more about it but but each one is nuanced yeah yeah we do we already do it everywhere that you could do it international I guess we don't do frontier markets but we think we do in emerging markets we do it in international develop US Canada we we are depends on the strategy you know either at at or near capacity or not at all I mean our large cap strategy still has I don't want I don't want it sound like too much of a crusader here we still have we still have a lot of capacity because it's large cap it's all relative though it couldn't couldn't accommodate 30 billion which you know most large cap strategies probably could so so we already do do it everywhere yeah yeah yeah well someday so I think we're out of time for this session if you guys have an itch of us and helped agree feel free to go up and no thanks I'd thank you

Channel: Quantopian

Views: 34,807

Rating: 4.9032259 out of 5

Keywords: finance, quantitative finance, risk, risk analysis, math, statistics, algorithms, algorithmic trading

Id: 3oVRjq50y_8

Length: 44min 36sec (2676 seconds)

Published: Mon Mar 06 2017