In this episode I am joined by Liqian Ren, Director of Modern Alpha at WisdomTree.
After receiving her degree in Computer Science, Liqian came to the United States to pursue her Masters in Economics. Liqian then did a quick stint at the Federal Bank of Chicago as an associate economist, before returning back to academia to pursue her PhD at the University of Chicago.
In 2007, Liqian joined Vanguard’s Investment Strategy Group, where she leveraged her background to perform economic and capital market forecasts, studies on asset allocation, and research into topics such as retirement income and investor behavior.
Liqian eventually transitioned to Vanguard’s Quantitative Equity group, where research efforts were focused on deep, stock-level signals analysis and portfolio construction. Becoming one of the first to act in a dual capacity research / portfolio manager role, Liqian developed a deep appreciation for implementation-aware research.
We spend much of our conversation talking about factors in both theory and practice. We hit subjects such as the risks of delayed implementation, mixed versus integrated portfolio construction, opportunities for factor timing, active versus indexed implementations, and how factors fit within a glide path.
Finally, we discuss Liqian’s new role at WisdomTree and new areas of research she is excited to pursue.
I hope you enjoy our conversation.
Corey Hoffstein 00:00
All right, all right. 321 And let’s go Hello and welcome everyone. I’m Corey Hoffstein. And this is flirting with models, the podcast that pulls back the curtain to discover the human factor behind the quantitative strategy.
Corey Hoffstein Is the co founder and chief investment officer of new found research due to industry regulations. He will not discuss any of newfound research as funds on this podcast. All opinions expressed by podcast participants are solely their own opinion and do not reflect the opinion of new found research. This podcast is for informational purposes only and should not be relied upon as a basis for investment decisions. Clients of newfound research may maintain positions and securities discussed in this podcast for more information is it think newfound.com.
Corey Hoffstein 00:51
In this episode, I am joined by Lee chin read director of modern alpha at Wisdom Tree. After receiving her degree in computer science, Li Chen came to the United States to pursue her master’s in economics. He then then did a quick stint at the Federal Reserve Bank of Chicago as an associate economist before returning back to academia to pursue her PhD at the University of Chicago. In 2007, Li Chen joined Vanguard’s Investment Strategy Group, where she leveraged her background to perform economic and capital market forecasts studies on asset allocation and research into topics such as retirement income and investor behavior. Lead Gen eventually transitioned to Vanguards quantitative equity group, where research efforts were focused on deep stock level signal analysis and portfolio construction. becoming one of the first to act in a dual capacity research slash Portfolio Manager role. Li Chen developed a deep appreciation for implementation aware research. We spend much of our conversation talking about factors in both theory and practice. We hit subjects such as the risks of delayed implementation, and mixed versus integrated portfolio construction opportunities for factor timing, active versus indexed implementations, and how factors might fit within a glide path. Finally, we discussed the Chen’s new role at Wisdom Tree in new areas of research she is excited to pursue. I hope you enjoy our conversation. Chen, thank you so much for joining me on the podcast today. Thank you. So we have a lot to cover here. And so we’re gonna quickly jump over your background and just dive right in. And I want to start in a place a question about perspective. So it’s been my experience that a lot of quants come into the field with a background sort of in mathematics, the hard sciences, financial engineering, data science, you actually come into the field with an economics background. And I was wondering how you think that has affected your perspective? And maybe how that perspective differs from other quants in the space?
Liqian Ren 03:07
Yes, actually, very much. So first, finance is a field grew out of economics, it’s actually a pretty young field. If you think about economics, as dismal science of 300 years old economics emphasize our decision making under constraints. So a good economic observation is where you observe people’s decision making process under limited and imperfect information, time and resources. So a good economic study, consider the human behavior biases, and its potential origins, where finance is very relevant, particularly in active quant. So when we say follow the money, it is really implying economic decisions consciously or subconsciously moving that money around, from which we could infer some economic or financial decision rules. So I think after living in China, where government has a very oversized control of personal decision making, and I found that a good economic training really makes me see things a little bit more clear, only an economist has the guts to kind of say, the misery of being exploited by capitalists is nothing compared to the misery of not being exploited at all. So
Corey Hoffstein 04:23
it’s definitely a true economist first.
Liqian Ren 04:26
Politically Correct. But economists, the world say no good financial economist will definitely look at the evidence and generally agree, at least given the evidence in existence.
Corey Hoffstein 04:39
Do you think that your background in economics has given you maybe your unique perspective around things like risk premia and investment styles as compared to someone who maybe comes at the problem more from a data science background that’s just going to try to look at the data alone?
Liqian Ren 04:58
Yes, because for me, and I think for WisdomTree as a whole as well, when we think about economic factor or a risk premium, we want to say whether it is based on sound economic principles. And I will give you an example. So for example, quality, quality factor in finance, there are many researchers which have argued, the quality factor can come from purely rational part, it’s a risk of premium based on rational behaviors, or that quality exists because of some biases. So, for me, I always start with, think about economic risk of humor, whether it’s grounded in economics. So for example, researchers who say companies that invest less, by the same time generate are higher profits should mean that they have higher expected return on capital, which means quality over longer time should deliver a value. So that’s a completely rational economic decision, or economic rationale. But you also have researchers say quality arose because people pay too much attention to the headline numbers. For example, when order numbers or look at financial news of a company, they just look at the earnings, how much is earnings per share, they the company beat earnings per share consensus, quality factor. And as you also know, the way it’s constructed, use a little bit more balance sheet items than just pure headline numbers. And that is you can argue that people don’t have time to go into detail and rely on facts or headline numbers. That’s a bias. And then by being more specific into the lower level items of the data, we can see some mispricing from the misbehavior. So for me, I think it’s really important if a factor is only based on pure statistical pattern, without a sound economic or human behavior, it’s really a little bit suspicious. So
Corey Hoffstein 07:03
we’re gonna dive into factors in depth a little later. But I do want to rewind a little bit prior to Wisdom Tree, you were actually at Vanguard, where you initially joined in their investment strategy group and focused more using your economic background on their economic and capital market forecast things like asset allocation, and actually published a number of studies on investor behavior. And I know that one of your areas of exploration was glide paths. Glide paths are an area that I find very fascinating. I think there’s a lot of innovative work that I’ve seen published about glide paths. And glide paths really have this huge potential impact on investor outcomes. But it does feel like to me at least, and this might just be a naive, outsider’s view that there hasn’t been a lot of innovation in actual implementation of glide paths. I know this is an area you’ve looked a lot at in the past, where do you think there is room for innovation and improvement in the world to glide paths?
Liqian Ren 08:04
Thank you. Actually, the first two strategies I worked on were asset allocation strategies, targeted funds, manager PR funds, which has a lot to do with saving for retirement and how to spend down in retirement. And I think it’s very relevant because there really two points. One is that currently, the retirement for one case base in some way is highly regulated. So for example, the infrastructure is not there. There are very few firms which offers some of the more innovative approaches like ETFs. I think like probably 90% of return plans don’t allow a person to invest in low cost ETF low costs activity, and ETF, as both of us agree has been really an innovation in the last 20 years, not just in tax. I mean tax advantage is a big part of ETF ETF, the trading of ETF is you can trade during the day, if you would have an ETF, it’s much more portable, if you have a mutual fund, the trading costs on TD or the other companies is close to $40, or retail investor. But a lot of ETFs is really commission free. Of course, we don’t want to encourage our people saving for retirement to trade unnecessary in timing, but what I’m seeing is some of the innovations in financial markets. Because the way the retirement landscape right now in the US wasn’t able to get into targeted funds. I think this area where things get a little bit move toward more innovations and the government to loosen up some of the restrictions that would have helped get innovation more into retirement. The second thing is really asset allocation. Because in asset management and you and I both know it’s very dominant in Dubai, a few companies in the place where the entry cost entry barrier is high. Then some of the innovation wasn’t able to get in. So for example, I personally believe in tactical asset allocation. I believe it just like factor timing, it’s a tactical asset allocation is very, very hard. But in some extreme situations, when valuation gets too high, for example, there are benefits of tactical asset allocation. And Professor Siegel, who recently have written a paper about tactical allocation across sectors as well. And it’s purely based on value. When a sector gets too expensive, you will think there is something which over long run that some kind of tactical allocation would help. So I personally agree, there’s some element of tactical allocation should be in glide paths. But of course, you competing against a very big companies again, so it’s gonna take a lot of education from you, from everyone to get people comfortable, or ordinary person really have to spend their time and resources to investigate if there’s tactical allocation, just like active pawns. So I think that’s another area I will say a second innovation. So the first is more kind of financial instruments that innovation like ETFs, and a little bit of telco as an option. The third area, which WisdomTree actually happened to have a lot of expertise in this is that we have a senior advisor, Professor Coughlin from MIT age lab, the way he think thought about retirement is not most of the bypass assumes most of the guide pass just assumes there’s one monolithic retirement, you retire, then that’s it, it’s almost like a you’re working, then that’s it without realizing that we could have a young no kids young with kids career, career, a retirement, it’s the same, particularly when a person has 2030 years of retirement horizon. So Professor Koffler mentioned that if you’re healthy, in that first stage of your retirement, you actually need a lot of income, because you want to go around the world and visit China or Spain. So all these requires you to have income without having to sell down your assets. Yet, because your retirement horizon is too high, you may not want to narrow down your equity allocation that fast. So all these I do believe that the are the innovations that will gradually get into the asset allocation, glide paths. And 10 or 20 years ago, a person’s access to a glide path is simple just through one target date funds. And that’s predominantly the only choice you have unless you have an expert doing it for you. But going forward, there are a proliferation of model portfolios. And that really was significantly change the way you could manage your money because it’s very costly to set up a fund, particularly a publicly traded fund, there’s as easy as so much regulations. But if you have a model portfolio, that you could sign yourself up on it, you won’t be committed to a particular fund. So that is another area I see more innovations will come folklife has, it also provides a competitive landscape, so that smaller shops who have knowledge and expertise in this area could have a shot, not at the model portfolio level.
Corey Hoffstein 13:31
One of the things that I find really fascinating about working at a large company is that a lot of large companies have an unbelievable amount of access to data and its proprietary data that they’ve collected over time. So for example, I think our friend Dan Egan at betterment, when you talk to him, they have been able to learn an incredible amount about investor behavior, through everything that they see in real time, over time about the investors that are on the betterman platform. And I know that when you were at Vanguard, you had access to a good amount of data about what Vanguard investors were doing as well. And so I want to tie this now starting to drop into factors as a whole, which I think is probably where we’ll spend the rest of the conversation. But there’s this big discrepancy in the factor research as to whether the explanation for factors are risk based or whether they’re behavioral. And I know you wrote a number of papers about sort of behavior and misbehavior of Vanguard investors. For example, you commented on very short term momentum effects that you saw, I wanted to get a sense from you, as you’ve sort of looked at that research over time and the research you’ve performed, how has it informed your views about factors in general and where they come from
Liqian Ren 14:53
definitely was a big firm you could access to a lot of information of their trading behavior. Are there actually I want to comment in the sense that actually, the more data you see, the more you realize that human behavior is so varied. There’s so much difference between people. I think that is something which I can never appreciate enough. I think as economist, you always think there’s a preference which is reflected in your utility function. And a lot of economics papers, really just assume one utility function for a whole population. But when you actually look at the behavior, you found that most likely it is because of the constraints. One person versus somebody, for example, females trade less than males. And it could be just females just don’t have enough time, or could be that they don’t have interest. So it’s very difficult to disentangle whether it’s internal, which I kind of think it’s rational, if somebody just don’t have interest, and then he doesn’t trade or somebody who is just too busy, and it’s constrained. So you don’t have the time to. So you will see the behaviors a difference. And then you will see that actually, when I think about the factors, I tried to think about this, as was that if you think that value or momentum works, is it from a sound person’s judgment. And for example, you can see that the momentum works, because there’s under reaction, which is behavior. But you can also says that momentan is just incomplete information, because it could be the URL, you know, the information, but I have not had time to learn about the information. So that is why my reaction is slightly lower. So I think both the economic decisions as humans make decisions with incomplete information with constraints of time, everybody have 24 hours, that is the biggest constraint on any human behavior. So if you believe these kinds of things, could lead to factors performing, I think, then you have a good rationale, because we are putting billions of money in making these factor beds. So when we make these factor beds, we better have a good rationale behind it a good human rationale behind it. Now, if on top of that, we also have a kind of a behavior buyers story. That is probably even better if person in the behavioral finance literature, there are so many stories where professors have put people through all these natural experiments to show that people’s behavior is not consistent. So inconsistency is definitely something it’s a bias without changing too much. You cannot say today, because this reason you like value, and tomorrow, you like value for the other reason. So humans without changing information, your behavior is inconsistent. With more facts, of course, you’re going to change your rationale. But if there’s no change of information, decision is inconsistent, then that is the misbehavior. And in all these behavior research, we have shown that humans indeed have a lot of behavior traps, we are loss aversion losing is painful for us, we have endowment a fact is that if we own something with you, it’s more valuable than if we don’t own it. Like if you own a cup, and ask you to assess how much a couple is worth it just by owning it, you will access a cup more value. And also, I don’t play poker, but a lot of poker players. If they lost something, they start to bet more aggressively. There is research because there’s so many poker rules. So you have data by these poker platforms to see whether people do that. And actually, another thing interesting is that Chicago Board of Trade our traders, if they have lost money in the morning, they are much more likely to take more risk in their afternoon trades, and they end up losing money overall. So all these it is really misbehavior. Human misbehavior. If you’re consistent, then the way you play poker, how consistently You bet, versus the way you trade should not depend on whether you just lost some money, but these are inconsistency which could exists and sometimes I will say augmented the fact that story.
Corey Hoffstein 19:29
So a couple of thoughts. First of all, I’m absolutely amazed that you’re allowed to work at Wisdom Tree with Jeremy Schwartz and not play poker. I’m sure he’ll try to fix that over time. We
Liqian Ren 19:39
did play poker a couple of times, I’m not very good. So I try not to participate as though I don’t lose all the money.
Corey Hoffstein 19:45
I’m with you. You would think that most quants know how to play poker. I think at least for me, the problem is I know I’m the sucker at the table. So I never want to play. I know I don’t know. I actually want to pause here and actually take a step back because I made an assumption that I want to correct which Just generally when we’re talking factors, I just assume everyone’s talking about the same factors. But maybe we can just real quick question for you. When you look at the factor landscape, what do you sort of think the big factors are that are potentially worth pursuing for investors?
Liqian Ren 20:17
I think, value momentum quality, momentum is a little bit harder for me, I think of value and quality. It’s very hard to pick it’s almost your kids value and quality, I think, has pretty high conviction, in the sense that, of course, there’s 50 different shades of value. So how you define value and quality, but I think in general, value and quality is well established momentan recently have you and I went to the Democratic critize quantum conference. And we have seen the paper like a 200 years of momentum, which has shown that momentum is also well, it could have 20 years of momentum not performing. And also momentum in some way requires so much of transaction cost that you wonder whether that is an area which I believe there’s some consensus that you could potentially extract momentum, but that’s an area where I pay attention to as well indeed, can you capture momentum, even factor exist? The other areas, I think growth a little bit the level? These are the kinds of factors by itself. Most of academic was a it’s not necessary, kind of stand alone factor unless you have strong belief that you want to time the factor. You like to own growth, or you like to own low vol. That by itself, is it a overwhelming factor or not? I think there are papers that still arguing. So these are areas were actually interested in research could potentially come because 10 years ago, Warren Buffet is being invested in using some idea of ROP quality related investment. But only recently quality become a kind of consensus factor. So on the
Corey Hoffstein 22:04
quality side do you think of because I sort of put factors typically into groups factors that tend to have that sort of monotonic increase in alpha. So as you have a higher Factor score, you get more excess return potential, and then factors that tend to be No, let’s just screen out the worst. This is really a factor that’s about minimizing risk exposure, value momentum, I typically put in the former camp, are these more Alpha factors and quality in the ladder? You mentioned quality a number of times now, do you think quality is a proactive, alpha seeking type factor that can be used by investors to potentially enhance returns?
Liqian Ren 22:42
For now? I do. I think one of the things I want people to pay attention is that all quantum research is not you just do it and then done, I think requires people to really be open to new ideas coming out. So for if you ask me today, yes, I believe quality is kind of as you can, it’s better used if it combines with value. So one previous idea is that he was made a joke of one of his friends, like he said his friend will buy those really suck stocks. It’s almost like the smoke, there’s to puff left and you smoke. But he’s not bad. He’s more about a good company, but still have a good valuation. I think if you combine with quality strategy with value that is similar to the way I think about I’m not just going to buy quality stocks, I also look at the price, I don’t want to buy a stock that is high quality, but it’s so expensive. So I think if you combined with value, it’s still kind of alpha seeking kind of a factor, at least for now. That’s the way I believe. Now, I do agree with you. I mean, if quality is still a little bit harder, in the sense that some of the features of quality, for example, is it too much risk is too much risk in your earnings growth. That sounds a lot like something you could use as a screening tool. I think it was a screening tool unless I see a very nonlinear relationship, like you said, you know, construction, a very nonlinear, I will use as a screening tool, but I think it’s so far, the traditional kind of quality measure is pretty linear.
Corey Hoffstein 24:24
So you brought it up. So we’re gonna go there, you talked about combining quality and value together, we’re gonna go to my one of my passion topics, which is multi factor investing and talking about the right way to combine signals. So there’s really two big camps right? There’s the mixed camp where you are building a multi factor portfolio by different sleeves. So you’ve got your value sleeve and your quality sleeve and then there’s the integrated camp, which is saying no, we really want to combine these signals and buy the securities with the best combined score and as with almost all things quant there are paid First supporting both sides published by very reputable firms. So I’ve seen papers published by Research Affiliates in AQR and Goldman Sachs, all making the complete opposite arguments with very convincing supporting data. But I know this is an area you’ve done a lot of research into and have some passionate thoughts on. So I would love to hear your thoughts as to maybe what you think the best way of building a multifactor portfolio is, and maybe the some of the core considerations that you think investors should take into account when doing so?
Liqian Ren 25:32
Yes. So I think, yes, this is definitely an area where we are very conscious. So I think the first approach usually is considered top down, you have different leaves, and then you combine this leaves putting money in differently, versus bottom up the way is you have different practice scores for each stock, and then you use certain weight, and then you come up with a multi factor composite score, and then use that score to construct your portfolio. So for now, I am still in the second camp. And I think from the economic point of view, the way I think about stocks is that in the sleeve approach, you could be buying a stock, which is very high value, but very low momentum, just because it’s in the value portfolio, or very high momentum, but very low value. And then just because it’s in the momentum portfolio, if you tilt into all those factors. In the bottom up approach, I’m looking at stocks, which in general, mostly, you know, average are good, a few characteristics are good. And there are one or two characteristics, which really made that stock stands out. That’s how you get a high multifactor composite. So I believe that way of constructing a portfolio at least the way I think of is having a team of stocks is I want generally not to bet that most of the characteristics and also stand out there kind of multifactor portfolio. Now I know that empirically, most of the empirical research I’ve seen has favored the bottom up approach. I would love to offline to read sleeve approach a little bit more. And I think, one I’d forgotten who wrote it, but one of the early argument was that in the sleeve approach, if you just lever up, then you can get close to the bottom up approach. That is from a few years ago, I don’t know whether those numbers are still with the new data coming in was this true? And other time, I will say that, even if it is that true that if you were to run a levered portfolio of sleeves approach that will be similar in terms of empirical numbers in terms of a bottom up approach. I will say that that’s not realistic, because most retail investors cannot level up. So that is the only piece I have seen. I don’t remember whom but the only kind of research I’ve seen, which says that the sleep the approach is not necessarily bad. I think most of the research Ising is about bottom half approach. So I think
Corey Hoffstein 28:09
actually on the ilmainen talked about that a little bit at the democratized quant conference, you mentioned where he was talking about that might be one of the limits to arbitrage and a diversified factor portfolio is that when you do the sleeve based approach, as you mentioned, you dampen the amount of exposure you have. And so even though your Sharpe ratio goes way up, your total excess expected return goes down through the combination. And so you actually would want to lever that portfolio because as you mentioned, because most people don’t, it ends up maybe not being as exploited as it could be. But it is really interesting. So let me get your thoughts on this one of the critiques I’ve often heard about a sort of integrated approach where you’re combining the signals. And this is a critique I’ve made in the past is that it invites much higher turnover because the signal itself is going to be mostly driven the variance in the signal is going to be driven by your highest turnover factor. So if you combine value and momentum say your portfolio is inherently going to sort of have to have the turnover profile of momentum
Liqian Ren 29:14
completely agree with you. That is why I think for multi factor you really have to pay attention to turnover which a lot of smart beta products, we have not started paying attention, but for our active strategies, we pay attention to turnover so that we could constrain the turnover and still try to capture the after cost because the reason you care about turnover is because transaction cost is high. So if you can find that kind of a little bit sweet spot, not completely 100% Unlimited turnover and still capture so they I do believe there is some work that can be done and that is for our international multifactor strategies. We have a turnover constraints.
Corey Hoffstein 30:01
So it’s totally unintentional. But it leads really well to where I wanted to go next. I know for you this is a comment you’ve made to me in the past, that you think that it’s really important for a pm to have this dual hatted perspective, both the quant research and the practical pm responsibilities, understanding the day to day implementation and operation of a portfolio. In your view, is that actually changed how you view quant research a little bit? Can you elaborate on that for me?
Liqian Ren 30:34
I’m so glad you asked. Because I felt like now I’m a little bit older, some of our my views definitely changed. So when I started it, I was really a pure signal researcher, I was more about can I construct the best signal and then according to this rule, that’s perfectly make my portfolio the research portfolio. That was like 12 years ago, I was still a little bit naive. And I think in some way is good because I was very focused on those areas. And then later on, when I start to manage the portfolio, I take on a little bit of portfolio management responsibilities on top of my researcher. And then I realized that this sounds so simple, but surely you care about the portfolio return more than anything else. Because how you come up with a signal go from signal to the final portfolio actually goes through so many things. And I love to use the word use the example. So because I think that’s the best way to illustrate how my view has changed. So for example, when we do act has momentum or value signal, we usually take end of the previous day value and give each stock a score. And then the portfolio manager is supposed to buy the stocks with the high school. But when I start managing the portfolio, I found that almost day to day, if you have cash flow every day, you realize that a lot of stocks, as soon as they opens, it’s up 10% Or it’s down 10%. Now your signal is calculated by based on yesterday’s, what do you do? And the traders will always say like, what should I do? And I was like, I don’t know what to do, because in my model, I never thought about this. So we actually it becomes so frequent, we decided to do a project to look that if you calculate your score by end of day yesterday, and the app is 10%. Is it still momentum stock or not? And then if it’s down 10%, does it become a value stock suddenly? So we actually specifically did a research project just to give us a sense of if this kind of thing happens, what to do. And I think if I were not managing the portfolio, I wouldn’t have thought that difficulty of capturing the alpha could be eroded by this kind of situation. Of course, you would think that looking back, you should have thought but we have limited knowledge and limited information. So that’s one good example. And the other good example is that transaction costs before when we think about the portfolio, we usually don’t really think that much about transaction costs. But small cap is very hard to trade and spread is very high. And you could buying a couple 100 shares, and it’s already three or 5% of the market. And do you want to do that, and the research in the research base they offer is very high in small cap. Alpha is very high in illiquidity stocks illiquid stocks, because there’s liquidity premium and all the other things, but can it be captured. So if you don’t do portfolio management, then you very much not appreciating how the portfolio construction can be very big part of the research. And I think if you look at academic papers, that’s where academic papers have not done a good job. There, they care about the signal so much. Yeah, they don’t make some even attempt to look at whether this signal, they keep saying that there is this anomaly. The anomaly is not an anomaly. If the trading cause kills it or because it’s too expensive to trade. If it’s a small stock, and not many portfolios hold it, then it’s not really an anomaly. So I think when you actually manage the portfolio, you appreciate that these kind of things way more than if you come up with signal and also another example is that we talk about sometimes how hard it is to construct a portfolio and how hard sometimes the data quality will play a role, or how hard sometimes you just don’t pay enough attention to some negative values or close to zero values. So sometimes a stock could be in bankruptcy protection and got a high quality score, mainly because they probably just got some Capital Infusion or some good quality earnings and all the write off some debt and from the balance sheet, just purely because the way we calculate the signal it could slip in. So some of the portfolio management tools will help us refine us the signal like sometimes, okay, when I tried to think about the quality, I want these kinds of companies to be quality, I don’t want the other kind of companies to be in quality. That is very hard if you don’t look at the portfolio every day.
Corey Hoffstein 35:31
So what was really eye opening for me in managing portfolios, as my career got going was recognizing that sometimes in the mentioned this a little that something like mutual fund flows or ETF flows, for example, there can be actual opportunities to use that potentially, to manage the portfolio in a better way. So as an example, I think there’s a lot of evidence that supports patient trading. But if you were to take a separately managed account and trade just a little bit every single day, that might be very frustrating for an investor not only to see all the transactions on their portfolio, but actually to see all the transaction costs, you are literally creating rebalance for them. But if you’re running a mutual fund, or an ETF, where there has to be cash coming in and out of the portfolio, regardless if you can use that cash, for example, to trade towards a portfolio you would rather be in. Well, now you’re sort of making lemonade out of lemons, I suppose, of being forced to trade. And I think those are other interesting realities of managing a portfolio that don’t necessarily come up in that one I actually think can be a potential benefit to a portfolio whereas I think most of these operational burdens tend to be more detriments.
Liqian Ren 36:39
Yeah, I completely agree. I mean, the crazy redemptions is a good time to sell some low scored stocks, because you have to redeem anyway, you have to come up with the cash. Yeah. So I think really, it gives the person a more holistic view. Because in the end, you create these signals, and you hope to work, the last step is actually implement in the portfolio implementation portfolio construction, could be done in a good way to minimize the cost to the portfolio. You’re trying to do it operationally, it is a challenge, because then you need somebody who understands the portfolio and also conveyed the kind of ideas to the traders. So operationally, it’s actually not easy.
Corey Hoffstein 37:24
So maybe tying in these ideas of operational realities into my next sort of line of thought and question here for you. I think it was really around 2011 to 2013, smart beta took off really in the market and continues to grow factor research really took off got adopted more quickly by institutions and individuals. At this point, it feels to me at least that factor research might be a little oversaturated, we see a lot of the same stuff. You’ve got now the factor zoo with hundreds and hundreds of different metrics. I think everyone sort of knows the pattern, find a metric, build some decile scores, test results, rinse, wash, repeat sort of situation. As you look at the landscape in the future, where do you see the future of factor investing going? Where do you think sort of the interesting edges of new research are going to be
Liqian Ren 38:15
I offers in the word quant in Chinese, the pronounciation. Corn in Chinese is a little bit similar to the word mining. And I think it’s a good way to think that way. It’s true that vector research kind of took off in the last couple years. But keep in mind, pharma wrote the paper way before that on value. Quality was not even a well known factor 10 years ago. So I think 10 years from today, our view of vector researcher will also change. So a few areas, for example, like affect the timing, I personally think it offers a little bit of value. It’s silly to leave those and right now, you will probably see a little bit more of that coming. Also growth factor. So growth factor, we all know that growth factor by itself didn’t deliver the value by when you combine it with other factors. It delivers value, but could it be that there’s some other ways of thinking defining growth, which it’s not necessarily high valued stocks? So I don’t know yet. But another area, which could be researched is size. So for example, I’m not yet able to go on Twitter, but I saw a tweet said as now he’s ended up where what is good paper on equal weighting versus cap weighting. For my one view, equal weighting is really kind of betting on size. Besides, of course, any factor or the factors we’re talking about is not pure factors. Size is correlated with value. So of course size in some way is betting on value as well. But we, so far, there’s not a very good story on why size is a premium. Like why just because smoking bunnies, maybe they are the behavior stories could be that smaller companies, not many people pay attention to. So big companies you get in the news a lot. But that’s a purely kind of behavior story. And I always feel like, if there’s a long term premium, I want to hear a more economic reason. So I think size premium could be a macro factor. That because size premium is such a long cycle, that and long cycle coincident with a macro kind of a factor. That right now, I don’t have an answer yet. But that’s a area where I think maybe in 10 years, we’ll get a little bit more clarity, the things we got the clarity really is built on 10 or 20 years of research. But I think the areas that we don’t we don’t really know sighs factor is, is it a balance sheet based factor? Or is my macro factor? Is it uncertainty based? Or is it really a long term factor at all, it has not worked for so many years. So I think that’s another area and also a little bit of cross asset. I recently saw a paper which talks about FX factors and bond risk premiums should be looked at together. Because all these effects is also related to all the interest rates related to foreign exchange, they I mean, economically, they should be connected. But it’s very hard to mesh these kinds of information together and estimate and find out the relationship or which causes wedge. So these kind of things, I do believe there will be more research coming out. These are some kind of a cross asset, that kind of back to research. So I think there are many exciting areas where if you want to go there’s two areas that will challenge people,
Corey Hoffstein 41:47
for my guesses, you triggered at least a couple of our listeners with ideas like factor timing, and the growth factor, which probably means they’re great areas of research, someone who’s going to be a guest this season on the podcast that I had, runs a very high concentration, five to 10 stock discretionary growth portfolio. And I was like, I have to have you on the podcast because I knew their track record. And to me as a quant, it’s such the antithesis of quant. And what was really amazing to me was how disciplined and quant like his process actually was. And I was sitting there in my head going, I think you could build a quant model out of this. So it’s interesting, right that those areas that probably people assume don’t work are probably the best areas of research going forward to your point. So but because you brought it up, we’re talking factor timing, because this is obviously a heated topic. So you seem at least tepidly pro factor timing. What are your thoughts?
Liqian Ren 42:39
Yeah, so I based on the empirical and also based on my economic idea, I think the more people do factor research and factor investment, you will think factor it itself also has its own momentum and valuation, some factors, we’ll get into the Fed and get out. So the idea is somewhat similar to stock. Now empirically, I do believe that the factor timing, the offer I can deliver is small, very limited, not even probably 1%. If you take into account all the transaction costs, and all the tracking errors is that you had to increase because the factor timing will deviate your portfolio even further away from CAP weighted index. So I think my view has always been Yes, I do believe it. Yes, I think it should be kind of well researched, but the amount people should expect it also should be small.
Corey Hoffstein 43:37
So I actually wrote a paper I think it was like probably four years ago that I submitted to a contest about factor timing, and they required 1000 word abstract and instead I wrote a limerick for my abstract. I just looked this up. The Limerick was when outperformance fixation leads to large inflow temptation, premiums erode, investors unload enabling factor rotation. I can’t believe I didn’t when
Liqian Ren 44:03
you should the way my English is not good enough to come up with the lyrics but I can definitely say that’s one of the best the lyrics.
Corey Hoffstein 44:12
I guess limericks are frowned upon in the academic circle. And it is interesting. There were some papers recently published that shows that time series momentum seems to work pretty well on factors. I’ve seen evidence that momentum works. Now it’s not post cost. And I think there’s always this complication of when we’re talking about factors to something you alluded to earlier, is it the long only implementation of a factor? Is that the long short is this really implementable? But it would seem that the same view believe that it’s behavioral biases are leading to the creation of these factors that the more the factors are adopted, you would expect those factors to potentially be subjected to the same behavioral biases.
Liqian Ren 44:51
Yes, I very much agree. And I think just like you mentioned about factory farming is that a lot of these research like you see it academic Paper. And it sounds very good. But you and I both know that as soon as we put all our constraints on top of existing multifactor, a lot of those does not add extra value. So every time there’s a new paper come out, of course, I always look up the ideas. But 70% of the time, academic research does not stands out, like it cannot really even be a multifactor. A simple multifactor. Again, when I think about the idea, even I myself on the factor rotation, I keep stressing is, the value you can add is a small moment. And similar, a lot of these papers, say by yourself, add so much value. But if you dig into the paper, it’s unlimited turnover. It’s small cap, it’s how you know, as soon as you put it in, in it’s not sector neutral, huge tracking error. So it’s long shot, where a lot of our portfolio is not long shot. So as soon as you put this in add on top of multifactor, how much those paper really has left over. It’s always if I have to make a bet any paper, somebody gave it to me, I’ll bet it doesn’t add value. So based on my experience as well, I’m always open, because I think that is where there’s a time investment, we have to investigate why it didn’t work out of one or two out of 10. It works. And that’s the happy days. So
Corey Hoffstein 46:34
So sticking with this theme of academic papers versus real world implementation, obviously, the proliferation of smart beta products has brought a lot of these factor ideas into investors portfolios. As you look at the landscape of available products that are out there to investors. I’m not asking you to name product names here, for sure. But are there ideas that you see out there that you go, wow, that was a really clever idea. I like the way that portfolio was implemented. Or conversely, you look at portfolios and say, Nope, I know from my research, that’s not such a great idea. Are there any portfolios you wish you had been behind and built? Yes,
Liqian Ren 47:14
definitely. And let’s not name names. But, for example, on both sides, some of the momentum portfolios, I really think, momentum and because of the nature of turnover required, you really want more frequently rebalance, but constrain the rebalance. To get to the momentum factor a little bit more on the good side is that I am actually I worked at the Federal Reserve Bank of Chicago for a year as an associate economist in 1999, to 2000, which is really interesting time for macroeconomics. Of course, for me, it was my first job coming to us and graduate, I always remember those days. At that time, I was passionate to be a macro economist, actually. But for a long while, I was also skeptical, I actually the Chicago activity index, I was one of the researchers who worked on that and stock Watson’s inflation, forecast and paper, you know, us at different macro zeros. So I worked on a lot of macro data, yet, for a long time, I was very skeptical of how much macro information can be used in factory search. But I think probably in the last couple years, I have started to think a little bit more in this area. And some of the implementation of taking into account all the macro information into portfolios, could potentially I’m not 100% convinced yet, but they’re really made me pay attention to it. So I think, kind of doing my own research and also looking at how others have done it. And really going back to the early years when I was just purely looking at macro forecasting, like inflation forecasting. So that’s an area which I think other there are some innovative strategies.
Corey Hoffstein 49:02
A lot of times when we’re talking about factors, it’s sort of pure factors in isolation. So just this idea of how can we outperform the market create excess returns, and the reality is, is typically down the line, an investor who’s trying to achieve some sort of outcome and ultimately is going to be implementing this exposure within a portfolio context. Going back to some of the work you had done prior to your work in factories, where you worked on capital market forecasting and asset allocation, and glide path work. I wanted to get your thoughts on how these ideas can marry. How do you think of factors fitting within the asset allocation? How do you think of factors from maybe a glide path perspective? Where do they fit? Should you be using the same mix of factors throughout an investor’s lifetime or is that maybe a younger investor has different exposures than an older investor How do you think investors should think about using these tools in their portfolios?
Liqian Ren 50:05
I would say that definitely. So for example, when you’re young, you can take some risk, then you should be going for much more for aiming for after costs outperformance when you’re young, you could take some risk. When you are older, and you have some needs in kind of income, then you probably would tilt a little bit towards kind of dividend paying stocks, dividend paying strategies, which gives you the income yet you don’t have to sell down your equity holdings. And also, when you are in more retirement age, you still want to invest in equity, but you don’t want to go for high volatility, even within the active portfolios. Within the active portfolios, we have active strategies, which is 670 5% tracking error, or you have strategies, which is much more simpler kind of dividend waiting earnings waiting a little bit of quality to load, those are lower tracking error ones, yet, sometimes it’s like the lower volatility ones, you’re still in equity. So I think when you’re older, I think one thing is clear is that a person needs substantial equity exposure, to get through the lifecycle. And the different stage, that kind of factors YouTube’s to, should reflect your age. So that’s how I think a little bit in terms of how a person’s like asset allocation, I think in terms of asset allocation, the one question is in the future sector mutual kind of multifactor strategy, would it become part of the core holdings from my point of view, a low cost, sector neutral multifactor strategy would potentially be somebody’s core holding, because that’s where you get a little bit factor exposure in every little way. And then based on tactical allocation, you tilted toward some factor based on a little bit of the market condition or based on your own house, because one thing we mentioned that prefers a confidence research is that if you’re not very healthy, then you cannot travel much, then you really don’t need that much income. You know, you stay around. And so traveling is a huge cause. But health care, because because a huge part of your other. So how do you pay for your house care it should you use some kind of insurance to cover healthcare visits, travel cost, which is much more flexible. If you’re not healthy, you need the constant house expense. Or if you don’t have as much money, you don’t have to travel this year, you can be a little bit flexible. So I think based on your needs, then the kind of factor util two is probably also going to be different.
Corey Hoffstein 53:01
Earlier this year, you made the move to Wisdom Tree, and you now have the title of director of modern alpha, what is modern Alpha?
Liqian Ren 53:10
Honestly, when I was given the title, I don’t know what that means, either. And when I joined John, oh, he’s the CEO of Wizardry, he’s at a you have the best title. Do you know that? And I know that I really know that thanks to all the nice people I met indeed, I have a great title, always great titles or cause greater responsibility. So I think Wasn’t she has pioneered this idea, which was that really in the marketplace, right now, you have the very low cost beta. And I mean beta in cap weighted the whole market beta, not 5g, but cap weighted 5g cap weighted is such a high active bet, that’s not low cost beta. That’s not beta. And then on the other side, you have active strategies, which charge 1% or more, yet, they are purely simple factor investing. So what WisdomTree is more than half idea is combining both of them. Still a big beta exposure deviates slightly differently tracking error based on the strategies range from 1% to five or 6%. And low cost active strategies, systematic, purely transparent. People know our holdings. We have lots of disclosure, on the strategies that are behind each of the product. So I think that’s really the essence of model alpha. And you’ll see that all a person should care about just like somebody who does research and realize portfolio management is important is that what you care is really about your off cost, portfolio return. So suppose there’s zero cost index, yet on the other hand, you can get 30 basis points motif After strategy, then if you do believe multifactor has value, then all I need is to be able to beat 30 basis points to add 30 basis points of value, then it will be better than a pure no cost or zero cost index. So I think this is where the idea of Manasa that you care about after cost. And we are going to do the systematic strategies, which is transparent and much, much less than 1%. In terms of cost. If somebody can consistently deliver value, like you mentioned, a very good growth Portfolio Manager or Warren Buffet, who have delivered value and can persistently deliver value, I don’t think there’s any problem they charge more than 1%. Because all you care is about is after the cost.
Corey Hoffstein 55:51
You mentioned the words systematic and transparent, which are sort of the hallmarks of the indexed quant space. And in a prior conversation, you conveyed to me that you really saw part of your role as being helping WisdomTree transition from a more index base approach to actual active approach which I think quants would sort of say, well, index base active, it’s all systematic rules based to me. How do you see the distinction? And why do you think it’s important?
Liqian Ren 56:25
I think the distinction in reality is actually less than most of the headlines suggests. So I went on a panel discussion, and the headline says, smart beta versus active quant, as if it’s a huge distinction, I think the actual distinction is smaller than people think. But the distinctions are still there. So for example, a lot of smart beta products have the rules, then they got hang up in the mentor. There’s no ongoing research. And I kind of view this a little bit differently. I still believe the process is systematic and rule based, but the process should be continuously improved. So for example, when we think about portfolio management, like when do you rebalance? Before, it’s all based on a calendar schedule. Doesn’t matter how much your signal has decayed? You will rebalance on that Friday. But I think when you do factor research, you you realize that if all you care about is factor exposure, then sometimes you don’t really need to turn over your portfolio. If your factor exposure is still very good after a month, or quarter, or you can turn over a little bit less. Or sometimes the market is very volatile, and the stocks move a lot, there’s a significant shift, you may have to turn over significantly. Now, it’s not that we just use our head and say how much you do turnover. It’s actually a systematic process. Because you look at the vector exposure, you have some kind of baseline rule, which makes you trigger whether you should think how much to turnover. So that process is still systematic, yet. It’s a continuous process, if we see that doing a ranking within sectors is not as good as ranking within industry, which for a lot of signals, it matters, how do you rent them within? If we see that and then through our research, and there’s good economic intuition behind it, then the process will be change it but the change of the process itself is also well researched. It’s also systematic. So I think that is where the difference will be. But that doesn’t mean that an active portfolio is going to be just purely because the portfolio manager just makes a decision, which I don’t make when I look at portfolio, I have some systematical rules based on how I make the decision of the portfolio. And that rule itself is also well researched. So that’s how I see the active quant as different from traditional, just small beta.
Corey Hoffstein 58:59
So what else is on your research agenda as you transition over to wisdom three that you’re really excited to dive into.
Liqian Ren 59:05
So when I moved to his injury, actually, sometimes me you look back and you look at your own blind spots. The two areas of research you actually have missed, I acknowledge. One is currency hedging. I’ve actually even researched some carry or currency hedging because they used to manage the global mineral portfolio. So currency hedging is very relevant particularly for international portfolio. But I have not devoted enough research on this. And then now because currently hygiene is such a big part of Western tree and now that I took my tongue in the study a lot of those and I found that I should have thought about their factors in effects as well this FX moment and FX valuation carry. So I think that’s an area where we are going to make a significant push in house as well. The other way I have also is to marrying the theme based strategies with investment strategy. So for example, not to get the brownie points with Jeremy, and I’m sure you know, he’s, he’s gonna be on your guests. But he has been thinking about this way more, I think in the frontier. So for example, platform based, it’s not just an idea. But it’s also has an economic rationale behind it, it’s of companies that has a distinctive advantage of just being the platform being the scale economy, because the word is so connected, you know, you and I would not have known each other. No, maybe we physically have met quickly. But I think we got to know each other much more by the online reading each other’s all these connections, made the relationship stronger. And then companies have these natural monopolistic kind of advantages, you hope these would be a company that overall would provide the longer term but in the olden days, I was a little bit skeptical of these seem based. Now that I think about it, I think it’s really interesting. But the key for me, I really think that is to marrying the theme, with the investment rationale, like I’m from China, and we have some strategies, which is talking about ESG strategies, we are so focused on E and S. But for companies like China, G is the governance structure is so much more important. So in some way the ACC state owned companies, you would hope that the governance is slightly better, more than the state owned companies. So I think you can think that ESG is state owned is a theme, but also needs to have an investment case behind it. So I think that is an area where I will learn and also explore much more. The third area is thinking more fixed income, active together with equity, and wisdom to be very lucky that the head of fixed income strategy, he does research together with us. He’s very knowledgeable in all things fixed income. And I can always pick his brain in thinking about how do I cross when I research about the equity, I will try to get his How do you look at the company’s their debts and whether there’s any information that could be kind of translated between bond and fixed income. So these are advantages, and these are kind of a little bit frontier for us. We don’t know whether you will get it but we actively researched as well.
Corey Hoffstein 1:02:41
What about the world of machine learning, that seems to be a hot topic in the land a quant these days, everyone that I’m talking to is at least dabbling their toes in the waters of machine learning how much is that making headway into your research process,
Liqian Ren 1:02:57
it is part of our research process, I have not found kind of significantly eye opening results yet. And I think machine learning will people will realize that look, simple regression is also machine learning, you can think that the learning it is statistical by nature, and also it is a nonlinear relationship. So there are shops, which have done a lot of research in this area. And they’ve have used machine learning effectively, we also want you and we active research in this space as well. This is an area where it’s unavoidable. You mentioned about the linear versus nonlinear, if you believe that it’s nonlinear, then when you construct your portfolio, you should be nonlinear as well. So machine learning gives you some of the capabilities to do that. But it’s very, very hard in stock selection. It’s harder than most people think. And I think some of the existing strategy, you can see their performance is not that good. So I it’s definitely every shop is going to devote some time to it. But we are very far away from putting 90% of your money in a machine learning algorithm portfolio. And one thing interesting is that after I moved here with a jury, I found that actually, the sales team have used machine learning in reaching out to advisors, and they’ve made great success. So machine learning sometimes helps in so many ways that the unexpected. It could be that it’s just in those relationship management areas, machine learning is much better, and then know how to marry that into Portfolio Management. Many of us do you have to think that through.
Corey Hoffstein 1:04:44
Yeah, it’s really interesting. A lot of folks that I’ve spoken to and this sort of rings true for me have sort of picked and selected different ideas from machine learning, machine learning, cross validation, testing, and a lot of that sort of stuff where it might be more On the robustness testing side of the equation than necessarily the feature selection, or have adopted some sort of ensemble methodology, which is, there’s been a lot of machine learning research into, but not necessarily gone as far as, hey, we’re going to train a full deep neural network to try to just throw some data and exploit these nonlinear features. And with varying degrees of success, I think a lot of people who pick and choose have certainly incorporated machine learning. But if you just sort of go all in and say that machine learning is going to be a better tool to identify signals, I think a lot of people have been looking at the same data for decades and decades now, it’s going to be a difficult challenge for machine learning to identify in that same data, necessarily some very nonlinear patterns that haven’t necessarily been seen before.
Liqian Ren 1:05:47
Yeah, I completely agree. It’s an area because it’s new technology and shows value, it changed the Alpha Go and be so human. So there definitely is something in there. But like any technology, electricity was invented a long time ago, and video invention, and the time we could have electronic recording podcast is years behind. So there’ll be so many trial and errors on there. So there’s something in there. But how do we use this technology? It’s not the consensus.
Corey Hoffstein 1:06:21
Last question for you. This is the last question of the season for every guest. And the question is this, the situation is that you have to sell all of your investable assets, you sell everything you own, that’s invested. And you can only make one investment for the rest of your life. Now that can be an asset class, it could be an individual security, it could be an investment strategy. So you could say I’m gonna buy a 6040 portfolio, for example, what would it be, and why?
Liqian Ren 1:06:50
Oh, by I think multi factor, multi factor kind of strategy, probably 1060 40 as an allocation, I think the thing for me is that, if you believe us the beginning of the hill, if people can come together and solve the problems, then there are still good potentials, that we could make some productivity breakthrough. If you look at the long the economic, the real economic growth has been 1.8% 2%, the last couple 100 years, if we can just make a little bit closer to that, instead of the 1.4%. productive work grows, if people more productive, you still believe us should have confidence as a country, the political, the economic, then you will think equity will still deliver value. Of course, I would like to have some fixing, going exposure just to against low risk. I think if I will much richer, I’ll go for 100% equity, but I’m not so and I still believe if you believe multifactor is based on human behavior, and humans still live than humans is gonna still have all these faults or these constraints, we only live certain 24 hours time is limited. So if you consider this then a multi factor should deliver value. So I think that is easy. But of course that is my because I’m constrained. But if you are very rich, they could invest in a new technology could really change the trajectory. I think that a lot of those grand things that didn’t come in but given my limited investment and my responsibility to my kids. I think multifactor boasting equity and bond portfolio in the public equity market probably suits me the best
Corey Hoffstein 1:08:44
return. This has been a lot of fun. Thank you for joining me.
Liqian Ren 1:08:46
Thank you. Thank you