Podcast: Play in new window | Download | Embed
Subscribe: RSS
“We’ve got a platform that we built along with Bambu and Apex Clearing that’s called Tango that has full robo-advice capability. The difference between a traditional robo-advisor and Tango is that it’s built using the GOE algorithm. So you may be an RIA firm that wants to expand, but needs the technology to scale that just doesn’t have the capacity to deal with a huge influx of accounts. But bring in something like Tango and offer it to a wide array of clients, then mine the accounts for when they’ve hit a threshold where it might be worth the conversation to graduate that client into a full service relationship.”
— Matt Macomber, VP of Digital Wealth in the Americas, Franklin Templeton
The WealthTech Today podcast features interviews, news, and analysis on the trends and best practices in wealth and technology for wealth management, asset management, and related areas. This episode is part of our December focus on compliance. We’re talking to influential industry leaders who can provide technology solutions that help advisors build stronger relationships, improve outcomes, and enrich their clients’ lives. A quick shoutout to our sponsor, the Invest in Others Foundation, please go to InvestInOthers.org, and be sure to subscribe to our show wherever you listen to podcasts so you don’t miss future episodes.
Companies Mentioned
- AdvisorEngine [22:50]
- APEX Clearing [16:03]
- Bambu [16:00]
- Coherent [27:25]
- Quantifeed [22:30]
Topics Mentioned
- Dynamic Personalization vs. “Hyper Personalization”
- Integrating GOE into Your Tech Stack
- Closing the Advice Gap
- Connecting GOE and AdvisorEngine
Episode Transcript
Craig: I’m so happy to have our next guests on the program, they’re both from Franklin Templeton, we have a Deepak Agarwal, VP of Global FinTech. Hey Deepak.
Deepak: Hi Craig. How are you?
Craig: Fantastic man. And we also have Matt Macomber, VP of Digital Wealth in the Americas.
Matt: Craig, great to be with you. Thanks for having us.
Craig: Nice to have both y’all, I like having a crowd. It’s almost like a panel where we have more discussion and then go back and forth. And I’m excited I’m actually recording this live at the Market Council Summit in Miami Beach. Where are you guys calling in from?
Deepak: So we both are based out of Bay Area in California and it’s nice weather here. So hope you are enjoying Miami Beach as well.
What is the Goals Optimization Engine?
Craig: Finally I have nice weather, I had to fly to Miami to get it leaving crappy cold New Jersey. So here we are all enjoying ourselves. So let’s just let’s just jump right into this. So Deepak, can you give us the 30-second elevator
pitch for the goals optimization engine that Franklin Templeton has released and that technology? We’re really interested in how it works and how enterprise wealth management firms can use it. So can you explain what is the goals optimization engine?
Deepak: Absolutely, great. So the goals optimization engine is an investment advice methodology that we developed in house and this methodology provides personalized portfolios. We are using dynamic program optimization as the underlying methodology. And the objective of this is to maximize probability of achieving any stated goal for individuals.
Deepak: I must mention that this methodology helps support both accumulation and decumulation goes across multiple accounts. The sole idea of this engine is to focus on probability of achieving a goal and that drives the asset allocation. The portfolios are created in a manner where these are unique glide path or investment path for each individual’s and lastly, this methodology can accommodate and be responsive for changing asset allocation depending upon the market movements, change in capital market expectation as well as good changes for the individuals.
Craig: So one of the reasons why I wanted to have you guys on as they think the goals optimization engine or GOE right, the GOE, the GOE application, I think it’s unique and it’s got some interesting features that I wanted to bring out. And one thing you just mentioned, talking about the dynamic programming optimization and how it changes it. How is it different because everyone uses glide paths, everyone’s got that, every financial planning app has that. So what’s wrong with
that and why does a dynamic asset allocation deliver better results?
Deepak: I’m glad you asked that because many times, the challenge is that we do a good job as maybe wealth managers and advisors in ascertaining what is the risk profile for an individual and then provide them a standard glide path based on their risk appetite or risk capacity, but then leave it to them with respect to the market movements versus following a more institutional approach or for liability driven. So we try to differentiate with goals optimization engine, where at every point in time, the probability of achieving the stated goal remains at the center. And this methodology dials up or down the portfolio risk with an aim to say what are the opportunities in terms of taking extra risk or reducing the risk so as the individual can meet it, and more importantly, it does not have any standard way of assigning an individual to a specific glide path, but it actually creates a unique investment path for each individual. And not only at the beginning, but at a regular frequency, helps in terms of assessing whether they are on track or they should be making certain changes in the portfolio.
Matt: And Craig, I would add to that efficient portfolios don’t always equal alignment with an investor’s objective. And really what we’re trying to think about, as Deepak said it’s kind of looking at how a pension fund might manage the liability looking at redefining risk, as a probability of not meeting a goal versus simply standard deviation of a portfolio. So really being investor centered, not just per capita market centered.
Dynamic Personalization vs. “Hyper Personalization”
Craig: And that’s one of the other things I like about GOE, as a consultant, we hear lots of buzzwords, lots of vendors are coming in with buzzwords, and one of them that I really hate is “hyper personalization”. Everyone talks about how it’s hyper personalized, but it really isn’t. It’s no different than what they had before, they just put a new name on it. And that’s why a lot of investment products aren’t really personalized. They may change a couple things, tweak something here and there. They do it once when they start and then they and forget it and let it just run. Maybe they check it once
a year. But they don’t really do a lot of analysis. It’s not very dynamic. So that’s why I thought that your dynamic personalization is really hyper personalization, because it’s evaluating each individual investor on their own, and the algorithm is doing the work that advisors can’t do in his head.
Deepak: Is absolutely Craig and in fact, I would add to that, from an individual’s perspective, when you are at a certain size of your portfolio, you have an advisor who is ready to support you from the advice perspective and looking at all of this. But the moment we go down on the chain, we know from our experience, that advice gap is a big challenge. And how this technology helps us is that we have looked at a scientific way of saying how do you aim for each individual’s liability via their goals and then use the technology to offer them an advice which is well suited to help them achieve their goals. And many times it’s not about the superlative outcome just in terms of performance, but it’s about saying how are you managing the risk and the return which is required for the individual and that’s where I we believe that this methodology is truly differentiated.
Matt: And further accounting for inevitable changes in investor behavior, life circumstances, withdrawals, capital infusions, and looking at the market. So if the markets are great, we adjust but do we always adjust when life circumstances change? I think Deepak brings up a great point around this idea of an advice gap and Craig, another buzzword that I don’t particularly care for, this notion of “at scale”, another one that’s over used. We truly believe that something like this where you said the algorithm is doing the work that we truly can look to address this, this advice gaps that we are seeing and really pad democratize access to some of the smartest or portfolio planning mechanism out there.
Integrating GOE into Your Tech Stack
Craig: You’re incorporating thousands of hours of research and lots of people’s smarts into the software. So how would an enterprise wealth management firm implement this product? I can see on RIA would use it, but how would an enterprise wealth manager from take the goals optimization integrate it into their existing tech stack?
Matt: If you look at this idea, we talk about a couple of things, platformization and componentization, these enterprise firms that have platform-ized, if I can make up a word and have an existing platform and tech stack and what we’re coming in with is this component, the goals optimization engine that we think is, in fact, a better mousetrap. But we know that nobody’s interested in leaving their platform and alt-tabbing out to another website, logging in, doing their work there and then having to go back.
Matt: So what we’ve built with GOE is for someone like an enterprise an entirely API-driven interface that will plug into any existing platform. Where most of these firms have planning software, have risk analytic software, and we’re not going to go in and replace those necessarily, but what we can do is take the inputs from those systems, hit the GOE engine, and then we turn this personalized portfolio recommendation and furthermore integrate those portfolios within their existing portfolio diagnostic tool. So we know that ease of access is paramount if this is going to be successful. And if we don’t make it simple and make the advisor and the home office offices life easier, then ultimately this is going to fail. So we put a lot into building out the API infrastructure that they will tie in seamlessly to an existing tech stack.
Deepak: In fact, in addition to what Matt mentioned about flexibility and customization, I must also mention that on an enterprise perspective this methodology has been developed in partnership with the academic rigor. So our team at Franklin Templeton partnered with two professors at Santa Clara University and jointly they have developed this methodology, and we also have received a process patent for this. So this all brings a credibility around what an enterprise can look for, as a solution. And one more thing which is very important from an enterprise perspective that we feel in our experience that they like to have their preferences and customization around some of the investment requirements. So this methodology can accommodate the requirement on asset class preferences, what should go as underlying fulfillment product as well as if there are certain capital market expectations of their own. It can accommodate that as well.
Craig: So can you give me an example of some customization that can be done with the goals optimization engine in that case?
Deepak: Yeah, I’ll add one and then Matt, may also have a few. So for example, in the defined contribution space, we are looking at a use case of GOE to offer managed account solution, and as you can imagine that in a defined contribution space you have a plan menu and even managed accounts solution needs to be offered. Generally it is preferred that it picks up the fulfillment from the underlying plan menu options, and we have been successfully working with our advisory firm partners to look at the plan menu options create the model portfolios, which suit to those specific requirements, and also accommodating to the capital market expectations or asset class. preferences of those large advisory partner firms.
Craig: That’s another advantage of the goals optimization engine is your customized solutions can handle different investment products and different capital markets assumptions. It’s not just one size fits all.
Deepak: Absolutely.
Matt: I think that’s right and you brought up an enterprise using this but the clients that we’re working with now, we’ve got an RIA firms, and we run the gamut of kind of interest in GOE there are some smaller RIA firms that are they’re more than happy to use Franklin’s capital market expectations, use our suite of 16 portfolios that are along the efficient frontier for GOE and so the underlying asset allocation product selection but also have the flexibility, and again, working with our Franklin Templeton investment solutions group, to bring their own capital market assumptions, bring their own portfolios, and have that level of flexibility. Let’s let the API and the algorithm do the work and use their investment methodology to integrate. We want to be able to accommodate folks who want to bring their own or folks that really want us to do the heavy lifting.
Closing the Advice Gap
Craig: One thing we were discussing before we started talking was the advice gap. Can you explain what the advice gap is and how go helps to close that?
Deepak: So that is in fact one of our four key themes within our group that we are very much focused on. Looking at as part of solving we are digital, and some of the unique investment IP. So the way we look at Craig from our perspective is that advice gap is where you have a large population, which is currently not being catered to by the traditional advisors via the traditional advice approaches. And the simple reason is that the account balance is not enough to justify the cost involved. So where we see the opportunity is can we create simple interfaces and this kind of methodology. We have GOE which can take certain inputs via that simple interface for each individual and then run the methodology via API’s to provide specific recommendations that can help in terms of identifying what should be the asset allocation, what should
be the investment portfolio for each individual and not only do that at one time on enrollment, but continuously monitor it based on changes in the capital market expectation or the topsy turvy movement in market as well as any changes in the goals or preferences for each individual. Now, while we are using technology, we can get it to a scale where it can be offered to a mass affluent and a mass market population, and that’s very exciting to us.
Matt: I would say that we’ve got clients right now who are maybe a decent size, good size RIA firm that is looking, like Deepak said, to expand into a massive segment that might not have been as appealing to them in the past. But we’ve integrated GOE for example, we’ve got a platform that we built along with Bambu and Apex Clearing that’s called Tango that is a full robo advice capability. The difference between a traditional robo advisor and Tango is that this it’s actually plumbed with the GOE algorithm. So you may be an RIA firm that wants to expand, but again needs the technology to scale that just doesn’t have the capacity to deal with a huge influx of accounts but can bring in something like a Tango and offer it to a wider array of clients then actually mine the Tango accounts for when they’ve hit a threshold where it might be worth the conversation to kind of graduate that client into a full service relationship. So we really think that you’re democratizing access to your smart portfolio recommendation and financial planning through GOE is a huge opportunity for this application.
Craig: Yeah, a lot of firms don’t know how to what to do with their small accounts. They all talk about the ability to get a lottery ticket, like bring all the small accounts in and hopefully a certain percentage of them will hit whatever your minimum threshold is for advisors, but with your software they can automate that whole process rather than having to check it on a regular basis. Oh, we missed this accounts been sitting for two years and meanwhile your software should alert them much earlier than that, right?
Matt: That’s right. And a great experience for their clients as well because they’re put on a not just a glide path but a glide path that’s really driven by again, maximizing the probability of achieving the outcome of their goal and it’s not a set it and forget it or target a fund. It really is this goal based framework that on a periodic, typically an annual, basis is going to get reallocated and reviewed to ensure that they’re on the best the best path possible.
Craig: So something you mentioned earlier was how Tango, the robo advisor which is built by Bambu, the company based out of Singapore, the B2B platform provider, built Tango using the goals optimization engine and they built it completely with API, it’s all API based. So can you talk about your API structure, your infrastructure, why you built it this way, and is it truly API or is it just some like integration magic in the backend?
Deepak: Absolutely. So in fact, we take pride in saying that we have used the latest technology to support this investment methodology. And you mentioned API’s is this methodology and this solution is now hosted, it is delivered via API’s, and it’s a bi-directional API, because unlike some of the traditional investment advice where you do the risk profiling one time, and then you allocate the client into one of the five model portfolios, and then let that be managed via standard life. Here we are doing or dynamic asset allocation. We are constantly monitoring different data points. So these API’s really helped us in terms of offering this as a solution behind the scenes, which can work with any platform provider, or any web technology or other firms who are ready to accept the solution via API’s in terms of an input. And to that extent, the infrastructure is really scalable. And I also would like to mention is that the way that we have developed this it is not limited to the US as a market, but can be portable across the globe in terms of different requirements on the portfolio, as well as on the advice methodology. So we are trying to look at different use cases across international markets as well.
Craig: So you got ahead of me and Deepak, I was goning to ask you a bunch of questions and you answered them all. I was gonna ask you, is this API bi directional or is it just like a bunch of polls, or a bunch of little pushes? But no, you said it’s bi-directional. And is it a large portion of the data or it’s just a couple fields that we can pass between applications?
Deepak: The way it works is that we just need five data inputs for each individual, and those five inputs will tell the engine to provide an output back and that output could be that you are doing well with the expected probability. So no need to worry, but I can assure you that that’s not the output for the majority of us. So some people are underfunded, some are over funded. So then the engine what it does is that it provides the output in the form of a recommendation of saying that to meet certain desired probability, you need to either increase your allocation or maybe defer your goal or some recommendation. Or if you are over funded, it can actually tell you that you know what you’re doing good so why don’t you dial up your risk, or maybe have additional goals that you can meet with this?
Craig: That would be nice. Here’s the question, when you build your own stuff, your own internal development, are you using your own API’s or is that just for show, just for external partner? We call it eating your own dog food. Do you actually use these API’s internally?
Deepak: Yes, absolutely. Because that’s something which we wanted to be sure of that it works internally well, and we can give some example that yes, we are ready to integrate with a third party. So I would like to mention, a couple of examples here. So one, we have our ultra high net worth and high net worth fiduciary trust business in the US and Canada. So we are currently working on deploying GOE for a use case, which can help on their client segment. So that is where I think it fits into our internal use case. And we are using the same API infrastructure to deploy that.
And secondly, what we also are doing is we for example, have strategic investment in forms like Bambu, Quantifeed, and some of the others in the US and international. So we are very much looking at making sure that the same API infrastructure is being deployed when we are looking at one build solution and last but not the least, our wealthtech platform form AdvisorEngine in the US. We are again using the same infrastructure and deploying GOE into their platform.
Connecting GOE and AdvisorEngine
Craig: That’s a big move. AdvisorEngine has been around for a long time, they were one of the first digital onboarding tools, and we looked at them for large broker dealers five years ago, they were one of the only players, so that was a great move by you guys to acquire them. So how do they take advantage of GOE like how would you connect GOE and AdvisorEngine?
Matt: Sure, so if you think about some of the use cases from an advisor perspective, and an advisor using something like an AdvisorEngine to come in and you’ve got your client information. And then step one, as a decision support system for that advisor around mapping to a specific portfolio. So going in putting the inputs might come from again, a planning software that you use, risk software might come from AdvisorEngine, hit the GOE engine and get that kind of initial portfolio recommendation. And then for being able to map the client to that portfolio and then manage it all through AdvisorEngine or interface through kind of the process of that customer lifecycle.
Matt: And additionally, a new feature that we’re working on is a digital portfolio construction tool and capability. So if you think about extending that use case to an advisor getting that portfolio recommendation in and actually being able to do a diagnostic decomposition of that portfolio, to do an X-ray on it to see what does that portfolio look like, you might have a client that’s coming with a coming with their own portfolio, as you typically would, load that into into a portfolio diagnostic, bring one of the GOE portfolio recommended portfolios and run a side by side comparison, ultimately resulting in a proposal that then you go out to that client as a prospect and say, look, here’s what we can do. And again, bringing that capability to advisor into to expand the reach of GOE and the ability to really help advisors kind of scale that business because Craig as you know, advisors would much prefer spending their time on client communications, building their business and and not on portfolio management. This move to serve advisors as financial life coaches, beyond portfolio managers, we really think we can help satisfy that and make that advisor’s life just considerably easier by doing the heavy lifting on not only that initial allocation, you get through an AdvisorEngine but beyond going sure management as for care and feeding of that client.
Craig: The care and feeding of your clients. Yeah. So how do you decide on these investments? That’s a big move, to invest in other companies, can we expect to see more tighter integrations and more features from those partnerships?
Deepak: Is it more about saying that how are we integrating within the FinTech firms that we have invested in or the third party platforms?
Craig: The ones you invest in like Bambu and Quantifeed was the other one.
Deepak: So Quantifeed. But yeah, so you’re right. What we’re doing is that, you know, from our group perspective, we are very focused on the strategic investment across these FinTech firms. So in some scenarios, we have acquired those capabilities like an AdvisorEngine, or the Financial Guard, which was acquired by Legg Mason in part that as part of the transaction, so we continue to look at how do we enhance the capabilities on these platforms, we are adding a goals optimization engine as well as adding some of the other digital capabilities. But besides that, well we have these strategic investment as I mentioned, so Bambu and Quantifeed are the firm’s which are focused more on B2B robo solutions, they are operating both in Asia as well as Bambu is operating in the US. And we also have minority strategic investment in Coherent, which is more of a insurtech firm.
Deepak: So to answer your question directly, yes, we are very focused on looking at different use cases. We are integrating with these and sometimes it may be a bunch of products like a Tango, where we have Bambu as the front end, GOE as the advice, and Apex as the custodian. But in the other scenarios, it could be where we are doing these specific deployments for large banks or maybe some of these platform where it could be a custom deployment, but it makes sure that there is a close connectivity between the capabilities these FinTech offer and what we can support the investment methodology as well as the model portfolio and investment products.
Goals Optimization Engine
Craig: We packed a lot of information into this one episode and we’ve hit the end of our allotted time. Deepak and Matt, I’m so glad to have you on the program. Can you tell us where people can learn more about the goals optimization engine?
Deepak: Absolutely, Craig. So in fact, we have a good set of content that we are currently hosting on our website. So that is one space where some of the basic information can be found out. And it is always our pleasure, both Matt and I will be very happy to provide more information and engage with any interested parties over here. In fact, that is part of our role that we are trying to promote. So for sure, we are very happy to support all of that and there are some white papers that are already being published as part of Journal of Investment Management and some other publications. So those could be a good source of information, but we are happy to share.
Matt: Further I would say TangoRobo.com will give a great overview of the Tango platform and and how the investment methodology of GOE is plugged into it. So encourage your listeners to check that out as well.
Craig: And also FranklinTempleton.com.
Deepak: Absolutely.
Craig: Matt and Deepak, thanks so much for being here.
Matt: Thanks, Craig. Always great talking to you.
