This is the phase when you can show what you as a Wealth Management Firm can do for your client. Whether you use Wealth Advisors with expertise gained from years of experience in the industry or Robo-advisors, APIs can collate data from many sources and AI algorithms can analyse and generate the best fit portfolio strategies for your clients.
This article is Part 3 of our series on the Benefits and Risks of APIs in Wealth Management (“WM”) and here we focus on the Investment Analysis and Advice phases.
In Part 1, we considered the benefits and risks of using APIs to enhance client engagement - https://blueconnector.co/blog/2020-06-11-wealth-management-benefits-and-risks-of-apis.
In Part 2, we focused on the Client Onboarding phase - https://www.blueconnector.co/blog/2020-06-22-wealth-management-benefits-and-risks-of-apis-and-rpa-client-onboarding/
Investment Analysis and Advice
The Investment Analysis and Advice process comprises:
- Understanding client’s profile, goals, needs, asset preference, and portfolio creation (“The Client”)
- Asset Selection
- Portfolio Allocation
Understanding The Client
The advisor builds a profile of the client’s current financial position: the client’s net worth, sources of income, expenses and responsibilities, and current investments by asset class. The advisor also needs to know the client’s tax status: in which geographies/jurisdictions is the client required to pay tax?
Next, the advisor seeks to understand the client’s investment time horizon and inheritance plans - when does the client expect to see a return? Under what circumstances would the client expect to draw down on their investments? What life events (e.g. marriage, birth, divorce, retirement) affect their financial needs in the near and long term? What is the client’s approach to generational planning, including arrangements such as succession in the family business, or in taking care of the family finances? The client’s vision for how their inheritance is allocated is vital.
The advisor would also seek to understand the client’s attitude to risk: how would the client feel if the portfolio declined in the short term? Studies have shown that people are more averse to the risk of loss than optimistic about the possibility of gain. To what extent is the client concerned about volatility of their portfolio and the possibilities of short-term losses? The advisor will want to build a risk profile of the client, comprising:
- Risk appetite (i.e. client’s willingness to take risk to attain the desired returns on the portfolio)
- Risk capacity (i.e. client’s ability to take risk to attain the desired returns for the investments)
- Risk tolerance (i.e. limits or boundaries of the risk or the degree of uncertainty the client is willing to take. For e.g. downside risk client is willing to withstand)
One of the critical areas the advisor would need to discuss is to understand where the client wants to and is comfortable investing: traditional assets, alternative assets, global pool, or domestic pool of assets. It would also be prudent to discuss regulatory limits or constraints, if any, on the client’s ability to invest in certain asset classes due to their country of citizenship. Lastly, to create a portfolio of assets the advisor needs to understand the risk return that the client desires; therefore, some considerations to understand -
- Does the client desire market return or above market return? This largely relates to traditional asset class mixture, discussed below
- Is the client most suited for highly liquid assets which are highly correlated to the market, or illiquid assets with low correlation to the market, or to private markets? Or a balanced mix of the two?
- Insurance considerations (Note: here we discuss insurance in relation to alternative assets. Other forms of insurance, e.g. life insurance, would be part of a full suite of wealth management and financial planning)
All this data can be collected directly using digital application forms from the client. Regulators typically require evidence that the advisor has informed the client about investment risks and the possibility of losses. Obtaining a digital signature of the client, agreeing that the information collected is complete and accurate, can provide this evidence, as well as assist the advisor in recommending the most appropriate investments. Data privacy, usage and storage disclosure to clients is important for ensuring compliance with the relevant data protection policy such as the Personal Data Protection Act (PDPA) in Singapore and General Data Protection Regulation (GDPR) in Europe.
Here we offer a high-level view of some of the options that the advisors might give investors to choose from per their risk profile and expected return. In the next section, we share examples of market data APIs.
- Traditional asset classes – equities, fixed income, cash / cash equivalents (money markets)
- Equities - stocks in publicly held companies (dividend, yield, or growth stocks, new issues (IPOs), or defensive stocks (not highly correlated to the stock market)). You can further split this category into common and preferred stock ownership.
- Fixed Income (FI) - sovereign bonds and corporate bonds are the most common type of FI investments. Generally, FI investments provide investors a steady stream of cash flow i.e. interest / dividend at regular intervals or at maturity. Investors in many countries can opt for Exchange Traded Funds and /or Mutual Funds as well. The latter allows the investor to benefit from a basket of securities rather than buying one company’s or one government’s bonds. Treasury Bills and Treasury Notes are also part of this category of investment. For the most part, FI investment is considered a conservative investment where investors should expect a steady return but not too high unless it is invested in illiquid / junk bonds. If a company goes bankrupt, then expect a holder of a bond to be paid ahead of common stockholders.
- Cash / Cash equivalents - Investors can lend cash for short-term for borrowers, e.g. smaller companies, to use for day-to-day management.
- Alternative asset classes – This asset class can be generally harder to price due to limitation in price transparency and liquidity. Here are some of the alternative assets to consider; however, this is not an exhaustive list: investment in cryptocurrency, private placement investment in start-ups / other private companies, venture capital, real estate, tangible assets, commodities, futures, derivatives, hedge funds, ESG (Environment, Social, Governance) investing, and SWAG (Silver, Wine, Arts / Antiques, Gold).
Use of Application Programming Interface (APIs) in Traditional and Alternative Asset Classes
Market data for publicly traded securities can be obtained from service providers including Refinitiv, Quandl, Xignite, and Bloomberg. These service providers typically offer a web portal for subscribers to access data manually, as well as programmatically via APIs.
APIs can gather information on current and historical stock prices, dividends, adjustments, and splits for publicly traded stocks. Increasingly, datasets are being made available for hedge funds, mutual funds, money market funds, annuities, and structured products too.
Data service providers are also making available a wider range of Alternative Data from non-traditional publishers that can give insights into patterns of economic activity. Examples include car sales, company hiring activity, company spending and payment patterns, and FX transaction volumes.
The two main types of API are:
- Request/Response: Data is requested by sending a Request message to the service and is returned inside a Response message. This is the simplest to implement and can be useful when data is required infrequently or on demand.
- Event-Driven: Subscribing to a service results in data being delivered whenever the data changes. This is useful for data that changes frequently or at unpredictable intervals. Instead of repeatedly polling for the current value, the service ‘pushes’ data to the subscribing application as soon as it is available but does not spend time and bandwidth if there were no changes. An event-driven or streaming API provides updates in real-time (typically less than 100 milliseconds), using bandwidth and processing capacity more efficiently than the Request/Response method.
In Part 1 - https://blueconnector.co/blog/2020-06-11-wealth-management-benefits-and-risks-of-apis - of our series we gave an example of connecting to a simple stock market API to request intraday prices for a stock. This used the Request/Response format, which is the most common and most standardised form of API.
An application seeking to receive data from an event-driven API must implement a mechanism to ‘listen’ for events. Some market data providers publish sample code or Software Development Kits (SDKs) to help developers incorporate API calls into their applications; an example is https://finnhub.io/docs/api#webhook for notifications of company earnings announcements and real-time trades of stocks, foreign exchange or cryptocurrencies.
Both Request/Response and Event-Driven APIs incorporate security, requiring applications to authenticate before receiving data from the service.
Cryptocurrencies are digital currencies such as Bitcoin and Ethereum, based on blockchain technology. Since the invention of Bitcoin in 2008, thousands of cryptocurrencies have been created, and exchanges exist to facilitate trading. Market data providers include those listed above, as well as providers specialising in cryptocurrencies. Data can be obtained real-time via streaming service or as a database of historical market data.
Prices for precious metals such as gold, silver, and platinum can be obtained through the mainstream market data providers as well as specialist providers.
APIs exist to facilitate trading in fine wine. Liv-ex is an exchange linking fine wine merchants with traders. Their APIs https://www.liv-ex.com/library/ enable market participants to bring prices into their own systems so they can be displayed alongside other product information for a wine, use the Price APIs to find trading opportunities, and maintain watch lists based on their own or their customer’s preferences. Physical assets need to be delivered and stored securely. Wine has additional requirements as it is sensitive to light, humidity, and temperature. Exchanges such as Liv-ex offer delivery, storage, and stock management services, which can be requested via API.
Unlike fine wine, fine art is not produced in batches. Leonardo da Vinci didn’t knock out a job lot of 1,000 Mona Lisas. This uniqueness makes art and antiques hard to value. Art consultant and appraiser Alan Bamberger https://www.artbusiness.com/marketdata.html describes many other challenges in attempting to find accurate data about art market sales. Exacerbating this lack of price visibility, as Malcolm Gladwell notes in his Revisionist History podcast episode on art http://revisionisthistory.com/episodes/42-dragon-psychology-101, art galleries and museums don’t include the value of their vast fine art collections in their financial reports, even when challenged by the Financial Accounting Standards Board (FASB). While we didn’t find any art market data APIs, there are services such as https://www.artprice.com that provide art market data as a subscription.
Refinitiv provides data for ESG from thousands of companies and there are also several ESG rating agencies such as Sustainanalytics and MSCI Inc. Investors may be interested in investing in companies and projects that are for sustainable development purposes creating low carbon footprint for example.
Portfolio Allocation and AI in Investment Analysis
The objective of modern portfolio theory is to create a portfolio where investments fall as close as possible to the “Efficient Frontier”. The Efficient Frontier https://www.investopedia.com/terms/e/efficientfrontier.asp comprises investment portfolios that offer the highest expected return for a specific level of risk. Returns are dependent on the investment combinations that make up the portfolio. Successful optimisation of the return versus risk paradigm should place a portfolio along the efficient frontier line.
Artificial Intelligence (AI) and its subset Machine Learning (ML) can be used to inform portfolio selection and allocation. AI can categorise the client into a risk category and allocate portfolio for a client based on similar client profile or a suggestion that is best fit for the client.
AI can also be used to predict the future trend of securities in each asset class and make investment decisions.
In addition to providing market data to subscribers, APIs can be used to share data models derived from AI/Machine Learning. An example is Singapore-based Robo-advisor Bambu: https://developer.bambu.co. Bambu’s APIs enable companies to leverage its’ AI expertise and infrastructure:
- AutoML: Create, build, and train machine learning models specific to your business needs.
- Customer Segmentation: Divide (cluster) your customers into groups based on characteristics, interests, and traits.
- Portfolio Builder API: Create Model Portfolios along the efficient frontier and get a recommended product allocation of each model portfolio.
Finally, it would be prudent for the advisors to review the assets chosen in the portfolio to ascertain it aligns to the client’s risk profile and expected return.
Risk Management of APIs, AI and Third-Party Data
In Part 1 - https://blueconnector.co/blog/2020-06-11-wealth-management-benefits-and-risks-of-apis - of our series we gave examples of the risks that need to be considered in Client Engagement. In Part 2 - https://blueconnector.co/blog/2020-06-22-wealth-management-benefits-and-risks-of-apis-and-rpa-client-onboarding/ - we did the same for the Client Onboarding phase. Here we add to it:
There are fundamental operational risks that Wealth Managers need to consider, which we highlight here; however, it should be noted this is not an exhaustive list of operational risks concerning Investment Analysis and Advice phase:
- Many regulators around the world have charged companies for mis-selling. The Monetary Authority of Singapore (MAS) has put a paper on fairness and ethics of using AI, if advisors are using this technology to analyse data and make recommendations for investments. https://www.mas.gov.sg/publications/monographs-or-information-paper/2018/feat
- Most countries require advisors to pass exams and obtain a licence to give advice.
- Sales staff playing the role of an Advisor? Is there a conflict of interest?
- Front line staff compensation mix? Some regulators have tied revenue targets and variable compensation directly to inadequate advice or inappropriate products sold to clients.
- How well are Environmental, Social, and Governance risk managed? Recently, the MAS published a consultation paper on “Environmental Risk Management Guidelines” for Financial Institutions.
- Robo-advisors are not kept updated with the latest regulations and internal policies.
- Client approval not captured vis-à-vis advice provided, risks explained, and portfolio selected.
- AI models can become outdated, producing too many false positives and false negatives.
- AI algorithms are not verified or systematically tested.
- APIs increase efficiency and reduce manual errors of data collections, which reduces this risk.
- RPA can transfer data between legacy systems to CRM or new core systems and reduce manual errors of data transfer and create efficiency, reducing this risk.
- As noted earlier, risks of data protection and data privacy (e.g. PDPA, GDPR) not managed.
- Due diligence of Fund Managers who promise higher than market return not done, rendering investment exposed to Ponzi schemes.
APIs are an efficient way of gathering market data to inform investment analysis. Market data providers offer data via APIs on traditional and non-traditional assets where there are established exchanges, or where there is both liquidity and price transparency. The range of data available and their method of delivery continue to increase and evolve.
Increasingly, APIs also serve as a way of sharing AI models, enabling firms without their own AI capabilities the option of white-labelling services provided by others. Wealth Managers are then able to incorporate a “robo-advice” investment service for existing clients, or for customer segments they would not normally serve.
Whether accessing market data or complete AI models through APIs, Wealth Managers need to manage the risks, as they are responsible to both clients and regulators for the appropriateness and timeliness of their advice. A portfolio recommendation is prejudiced or is based on inaccurate or outdated data is dangerous for the client and can result in the Wealth Manager failing to meet their duty of care for their client’s financial wellbeing.
Wealth Managers need to understand the accuracy, timeliness, and provenance of the data on which they make client recommendations. The firm’s risk management function needs to concern itself with the end-to-end data lifecycle in meeting their client’s needs.
The next article in this series considers how APIs enable Order Execution in Wealth Management firms.