As Europe grapples with inflation, financial institutions are tirelessly searching for new ways of improving their business models. Whether updating their services with digital channels, creating more financial products, or exploring untapped markets, the financial executives carry on balancing compliance and innovation in their work.
If your organization is embarking on a bold expansion to new countries or updating their selection of investment products, it is critical to ensure that your financial analytics suite is attuned to your strategic roadmap. In these cases, it is imperative to check whether your financial planning software allows for detailed tax modelling of a target domicile and a swift process for mapping your new products to the underlying risk factors. A flexible and granular approach to financial analytics can help your customers reliably navigate the ever-changing financial landscape, securing your place among the vanguard of the industry.
Today we sat down with Lars Larsson, partner and quantitative finance expert at Kidbrooke®, to better understand how financial organizations can leverage the granularity of modelling and a detailed approach to representation of taxes to support their complex expansion roadmaps.
I worked at Goldman Sachs for ten years as a strategist for the investing and lending desk. I helped the desk hedge their positions based on macroeconomic, fundamental and technical factors. I was also responsible for the development of position risk reports, generating crash risk measures for the portfolio, pnl projections, conducting ad hoc scenario analyses, as well as suggesting tail risk hedging strategies.
I joined Kidbrooke® in 2015 and since then I have helped develop the simulation engine underpinning our financial planning software offering.
Risk factors can be thought of as asset classes or sub-asset classes, such as investment grade bond total returns or the total returns of an equity index etc. Conceptually, in the OutRank® - world, a risk factor can in turn build on several underlying model factors (for example, the total return of a bond index could depend on interest rates, credit spreads, default risk etc).
Very concretely, the manifestation of a risk factor is, for instance, a 5,000 scenarios x 984 months grid with each entry being the total return of the risk factor in that month, for that scenario. With these risk factors in hand, it’s straightforward to map the exposure of a financial product (such as a mutual fund or an ETF) and from there project its expected long-term future return. It is obviously possible for two or more different risk factors to be combined to closely mimic almost any kind of financial product. The end result is that you have, for instance, 5,000 different investment outcomes for your fund over a range of horizons. These outcomes can then be used to evaluate, for example, the feasibility of a savings or pension plan (feasibility = probability of reaching a specific savings goal) over a wide range of horizons. For any given range of inputs, our financial planning sofware generates a range of outputs.
A key part of the simulation is the dependency structure, ie. the mechanics of how the various factors interact. I don’t want to give away too much of the secret sauce here for obvious reasons, since this is essential to create a realistic projection into the future.
Essentially, financial risks are never considered in isolation, they impact a range of other aspects and financial products. Using a risk factor-based approach requires a macroeconomic view of a wide range of variables including, but not limited to, monetary policy, geopolitical developments, inflation, interest rates, currencies and economic growth trends.
We do so in close cooperation with our clients. The main objective is to represent their fund universe (or single stocks) as realistically as possible. Firstly, we break it down by addressing questions like “What is the key use case required right now?”, “Is the end goal a stand-alone investment customer journey or does the client want to simulate the end-customers’ entire balance sheet?”. The scenarios fuelling this method can easily incorporate wealth managers' outlooks on how economies will develop into the future (in the business, this is referred to as “house views”)
We structure the API and its endpoints to be easy to use and in line with our client’s expectations. Accordingly, we can develop new functionality based on our client’s needs within their financial planning sofware. It’s only because of our granular approach to the underlying risk factors that it is possible to accurately model the future market development of an institution's product universe. This is important because end-customers who receive advice accurately reflecting the effect of any variable will feel significantly more confident in the robustness of their decision making.
We release a new calibration of our simulation engine once a month, and one way to ensure that things are on track is to work out the change in the feasibility of a range of the client’s model portfolios over time. This means we look at their changing needs and requirements based on their end-customers’ demands. If there’s a big jump up or down in feasibility, we drill down into the simulations to ensure that the change is economically justified and not due to a miscalibration or bad data for example.
The main trick here is working out how to reuse existing risk factors as much as possible while still achieving a realistic representation of the new risk factor. For instance, if you’re looking at adding a new currency as a risk factor in your financial planning software, but you find out that the currency is pegged to a currency that you’ve already modelled (let’s take for example the United Arab Emirates Dirham, which is pegged to the USD, which in turn means that we don’t need to create standalone simulations for the Dirham, thus saving the client significant time and money).
Mapping the client’s product universe to our risk universe is the first step. OutRank® enables our clients to choose different levels of granularity when it comes to modelling their product universe and easily adjusts to the institutions' preferred risk factor universe. Achieving relative flexibility regarding the level of modelling granularity and implementing house views to their financial planning sofware has always been a priority for us.
In the next step we decide whether we need to add new risk factors. We then summarize our analysis in an annual validation report so our clients can form their own opinions regarding the reliability of the model.
The amount of time it takes is highly case-specific, for example, if there’s a particularly exotic product, it could take quite some time compared to a case where it’s another equity fund type we have already modelled.
In Sweden, it’s a very straightforward process in the sense that there’s an “official” tax calculator maintained by the tax authority that one can check the results against. This has proven immensely helpful to us because nailing tax laws can be tricky. We have also worked with some of our corporate partners to ensure that the more intricate parts of the tax system are correctly handled. We prioritize regulations and compliance and therefore examine the rules and regulations of international FCAs and constantly update the underlying processes and documentation as regulatory frameworks evolve.
On the order of a man-month. To maintain the model driving our financial planning software and ensure it’s up-to-date, we run overnight processes that compare our results to the official ones. This is of course a very reactive approach (more of a worst-case catch-all), so in conjunction with this, we also try to stay on top of new tax laws. It’s important to note that this is easily done in Sweden, given you can subscribe to all law updates via email. This way we stay well ahead of any changes.
We have added comprehensive tax logic to OutRank®’s simulation capabilities. This means that our clients get information about projected returns and learn about the tax implications of each simulated path. It is also possible to model any required tax domicile at request, so you would not need to be on the lookout for a new provider in the target country.
Apart from Sweden, we have completed the tax modelling process for England and Norway at a basic level and are looking to expand further in the near future.