vestr’s Head of Business Development, Stefan Wagner, speaks with Richard Waddington, Founder of Sherpa Funds Technology, about how to enhance investment returns through data-driven decision-making.

They also talk about the following in this episode:

• Optimal risk sizing in portfolio management
• Data driven decision-making
• Attributes to define the exposure of an investment portfolio while taking into account its limitations

and more!

‍

Tune in now and get the latest insights on structured products, technology and investment management:

Schedule a call with us for additional questions.

No time to listen? We summarised the podcast for you:

‍

Stefan: I will speak with Richard Waddington, CEO of Sherpa Funds Technology in Singapore, about optimal risk sizing and portfolio management. What is Sherpa Funds Technology, who do you help, and how do you achieve it?

Richard: Our raison d’être is helping people with data-driven decision-making. And that is quite a broad topic that is fairly well known in various industries. Investment management, particularly at the portfolio manager’s side rather than on the asset-selection side, data-driven decision making is actually quite rare. So, when we talk about ‘data-driven decision-making’, I deliberately take that phrase rather than ‘optimization’ or anything mathematical. Because it is not about having some fancy tool or some further math. Both of which we have of course. But it’s about helping a client with a process that helps them get better at what they are already quite good at. So, data-driven decision-making allows them to make decisions around how to build risk taking portfolios that makes the best use of their skills. 

‍

Is there an example of how you can show this since it is rather abstract?

Yes, an allegory is that it’s an investment portfolio. And really what I talk about is an institutional investment portfolio. Let’s take pension funds, mutual funds, rather than wealth products. When you talk to people who manage third-party money, a lot of them will talk about line item risk in their portfolio. So they will talk about Apple, Credit Suisse, or whatever they have in their portfolio. Because they are analysts at heart and they have years of experience in analyzing. This is my analogy; they are thinking of the portfolio as a salad, a plate on which lots of different things sit next to each other. They talk about “over here, we have the carrots, there we’ve got the beetroot.”

But when you invest, or any investor in a pooled assets investment product, you actually are buying a soup. What you care about is how the thing has been put together. Because the experience that you have as an investor is the experience of the whole thing at once. It doesn’t matter what happens to the beetroot or the carrot or whatever. What matters is that the whole salad has been put together properly. So the process of converting them from the salad to the soup is a process of understanding not only how the ingredients were originally chosen and what they are meant to do, it’s also what is the thing you’ve sold to the investor.

It’s about the investor’s expectation. About what you said up front to the investor that you’re going to do.

‍

Then, how do you determine the exposure? Do you do a factor analysis?

Different end-products have different attribute requirements. An attribute might be a factorness, it might be a barofactor, it might be some internal factor. Some clients may have very specific requirements to minimize, or maximize, or control particular attributes. But that is not the end goal. That is a constraint on the end goal. The end goal is the best expression of your view within those constraints. So you can say “I have a view these stocks are going really well, or these stocks are going really badly. And on top of that I have a requirement that my growth factor should be less than something and there must be this much of small caps.” Or whatever it might be. So there is a view, and then there is a set of requirements (constraints), and when you want to create that final product you need to express that product in the best way possible for the constraints that you have. And then you have a final weighing, a balance, you have to do. How much do you want to express your view and how much do you want to insure against concentration risk.

‍

I suspect that takes a bit of time for you to work out these attributes and constraints with the client?

Indeed, at Sherpa our engagement model is very front loaded in terms of effort on our end. The very first thing that we will do is shine a torch on the decision-making that is going on within the firm. Even that process, managers find that incredibly helpful. There are lots of things they haven’t actually formally written down.

‍

So, is there an iterative process over time to demonstrate your value?

The simple proof where we’re adding value is actually an easy part. We have clients who we’ve worked with for 3 or 4 years and we can show them the difference of month-on-month returns of our recommendations vs what they were doing before our recommendation. Overtime we keep all that data and we can show them. The demonstration of economic value is easy. The demonstration of process value is much softer. That is clear in the discussions we have with the PMs, where they request for an hour a month with Sherpa. During which they will just sit in and listen to what Sherpa has to say. It’s a very different interaction to a research provider. When a research provider talks about their own views, we talk about the clients’ views. 

‍

Do you also measure how much risk you are adding to those portfolios?

Everything in our view is about downside risk. It is not volatility, we don’t look at the linear standard deviation. We look at accelerating downside risk. If you lose 2%, it might be -1 unit of risk. If you lose 4% it could be -6 units of risk. With 10% it’s -50 units or risk. So, it heavily penalizes downside in our analytics. The point of what we actually do is creating expressions of risk of portfolios, which when the clients views are correct, they get the same return as what they would have done with their portfolio. When they’re wrong they lose much less. 

If the client is prepared to take an amount of risk we will tune our solution to have the same upside risk as you. But because we are doing the maths in a slightly better way, the downside risk in our portfolio will be less. 

Then over time you can see where on any given day when that portfolio makes +3, ours makes +2.9, +3.1, somewhere around there. When their portfolio is down -2, ours is always down -1. So, you can imagine that scattered graph, and over time you build up that scattered graph. Based on that we can explain to our clients that we are giving them as good as or better an expression of their view, when they are right. And we are controlling the downside much better.

‍

What does your client base look like?

The majority of our clients are dealing in equity trading, we have a couple of macro. We can talk about why that is one side vs the other. A lot of our equity guys, 80%, are not only benchmarked to big pension funds, big mutual funds. It is quite an involved process and you have to have the adequate resources, both financial and people-wise, to spend time with us.

‍

Is there any myth in your industry that you would like to debunk?

When a portfolio manager has to build a product, expecting them to do everything we have just been talking about in their heads, and to get a better than 50th percentile output, is impossible. It is not really a myth, it is just something that people do not really think about. If you are running a book of 60 assets, you are got a 60-dimensional problem with maybe 4, 5 different constraints, including benchmark, risk and sector. The idea any PM can do this and get anywhere near a good result is just a fallacy. 

Experienced PMs - and all of our clients have been managing money for 15, 20 years minimum, - solve this problem heuristically without real process. And they are never too far off the median result. That is, they are this experienced because they are good. 

The point is, a process can never get you to the very best result, you don’t know what’s going to happen in the future. But you can statistically go from the median 50th to the 65th or 70th percentile consistently by improving your process. 

So, the myth is: “It’s the only way to do it. We do not really know how good or bad the process is.” The reality is, we can tell you exactly how good or bad it is you’re doing. And we can bump you up into, rather than being in the median to being 60th 70th percentile, our experience shows us about 200 or 300 bips per unit leverage.

‍

Last question, what are your up to 3 favorite finance movies?

Risky Business!