By James Baty, Senior Vice President and Special Advisor at US Capital Global
It is vitally important to understand the undeniable primary principal of investing: “The method of investing is irrelevant compared to the value of the investment itself.” In considering where Artificial Intelligence may apply to Private Securities, let’s first consider the basic private security issuance process, and the relevant actor ecosystem, along the way noting some of the areas of recent FinTech innovation. We address what are ‘private securities’, how are they sold, what are the functions in that process and what are some key concerns that AI can address? Then we’ll examine three specific areas of possible related AI application in the near term. For illustration, the definitions and regulatory references apply to the US market, but are similar in other jurisdictions.
Private Securities – The Basics
What are the basic types of private securities?
The key ‘private securities’ are equity shares, fractional loans, and derivatives.
Equity shares – stock, or shares, are a main source of finance for a company. These convey fractional ownership to the investor, with typical owner rights… voting, dividends, potential capital appreciation, and liquidity options. Key concerns include the mechanisms used to value the company for pricing of the shares, and any market or operational risks. Recently this form of ownership investment has been augmented by the potential advantages of digital securities, i.e. ‘tokenisation’ that may enable transactional efficiencies and easier liquidity in a theoretical secondary market.
Fractional Loans – borrowing money is another key source of business capital, appropriate to operating businesses with assets and income. In contrast with single source loans from financial institutions, fractional loans are portions of loans, that are offered to individual investors. While equity shares offer potential appreciation, fractional loans offer a specified income stream (still of course with some risk). On the digital front, this vehicle has become popular with investors via the creation of many new peer-to-peer lending companies.
Derivatives are a financial security in the form of a contract where the value is derived from an underlying asset and its potential change in value or income stream. Derivatives include futures contracts, forwards, options, and swaps. While a significant portion of derivatives are used historically to hedge other investments, derivatives are also a way investors can participate in value appreciation of underlying assets, but don’t carry the ownership participation of equities. Technology for data streams of complex market information provides additional fintech innovation for derivatives contracts.
Whether the private security is an equity, fractional loan or derivative, there are risks and the investor must be concerned about valuation. Without yet considering applications of artificial intelligence, it is useful first to understand how these private securities are sold, the nature of their ‘regulation’, and what are the entities providing key functionality to support the private securities market.
Who can buy a private security?
Under the US Securities Act of 1933, a fund must either register its securities offering with the SEC or operate under an exemption from SEC registration. Under specific regulations, ‘Private securities’ are exempt from registration, thus less regulated than the public market, and as such are reserved to sophisticated investors, meaning those with sufficient investment capital to be able to assume the risks. In the US this basically includes ‘Accredited Investors’ or ‘Qualified Purchasers. The majority of unregistered securities offerings utilise the exemption under Rule 506(b) of Regulation D, which participation to ‘Accredited Investors’, basically those having either a net worth that exceeds $1,000,000, or income in excess of $200,000. Section 3(c)(7), exempts funds from registering if they are owned exclusively by ‘Qualified Purchasers’ who own $5,000,000 or more in investments. There are other regulatory nuances (e.g., limits on numbers of participants) and mechanisms of institutional participation, but these categories are the basic participants in private securities – those who have significant investment capital, and can assume the risks of evaluating unregistered securities.
How can private securities be advertised?
Rule 506(b) of Regulation D is restricted to private network solicitation, while Reg C allows general advertising. Regulation S is available only for offers and sales of securities outside the United States by US firms, but of course these may be subject to the foreign jurisdiction resident’s regulation. There has been a large amount of development of the digitisation of securities, however, as we can see from the press and the general response of global regulatory agencies, from the regulators perspective, while digitisation of securities may offer improvements in transactional efficiencies, from the regulators perspective, a digital security is still a security, subject to the same rules.
Who can custodian private securities?
Obviously the buyer themselves may hold the private security, or the security may be held for the buyer by the Issuer, or a regulated person under a power of attorney (POA). This includes commercial custodian firms. An interesting new fintech challenge includes the unique issues presented by custody of digital securities. While some promote the basic idea of ‘self custody’ by the investor based on possession of the digital key, there are also commercial custodian firms for digital securities.
How can secondary sales of private securities be done?
Basically secondary sales, must conform to the same rules and to the same persons as primary sales….except that to advertise a secondary sale one must have an additional license. Alternative Trading Systems (ATSs) which have been around for 50 years or so, enable access to these secondary trades. And there is a flurry of new activity around existing, and newly formed, ATSs enabled for the secondary trading of digital securities.
Private Placement investments offer sophisticated investors the ability to invest in equity, fractional loans and derivatives
How do you aggregate demand/participation and what use is an LP, SPV??
You can make a share a digital security sell and trade similar a public security, but there are limits to the number of participants or you are required to report like a public company. Another alternative is the formation of a Special Purpose Vehicle (SPV) which can sell a derivative …economic value in the pool of shares, which allows greater participation in income, while also avoiding cap chart complications of direct participation. You can do the same thing with a pool of loans and sell a derivative of that pool which avoids the 100 person limit for a fund. However, the use of SPVs raises additional information and compliance issues – it is very important to make sure the counter-parties are regulated and trusted by the manager of the SPV, understanding the valuation of the securities and trusting the platform where you sent your money to.
Private Securities – The Ecosystem
So what roles do securities professionals provide in this process?
We have all seen in the past few years where issuers have tried to navigate their own issues, and in some cases this resulted in regulatory citations, return of investors money, or loss of investments to fraud. The case of Theranos, in spite of the presence of lawyers, but without a licensed responsible broker dealer, ended in a lack of transparent valuation and a major fraud case. Countless ICOs tried to skirt the rules altogether, and in the best cases had to pay fines and return funds raised, while in the worst cases simply disappeared with investors funds.
Securities professionals, licensed professionals…. broker/dealer, attorneys, custodians, transfer agents, ATSs; each provide separate and specific functional roles – oversight, due diligence, transparent management, independent transfer and controls. Beyond the functional roles of these professionals, they also help construct an overall strategy for issuers that including deciding on the appropriate initial vehicle (e.g., 506C), the relevant contract terms (re. voting rights, right of first refusal on secondary sales [ROFR clause]). And in the longer term what is an appropriate multi-stage plan e.g., combining an initial round equity financing, with subsequent growth financing, operational loans, and subsequent liquidity events. Beyond the ‘fintech’ considerations, we can see that this involves important ‘regtech’ issues (e.g., understanding the value of general solicitation issues). For investors the ecosystems can help select appropriate individual investments and construct a portfolio that addresses their life goals and investment targets.
The Private Security Ecosystem
So we can see that, beyond the issuer, there are special defined roles for securities professionals that contribute to the success of issuers and investors. Broker/dealer, custodian, ATS, etc. and these have been augmented for digital securities by specialized security token platforms and traditional roles such as custodians and ATSs that are extended to accommodate digital securities. Generally, the broker/dealer acts as the ‘ringmaster’ calling on other professionals, e.g., a token platform, as appropriate.
Where Can Artificial Intelligence Help?
While there are many popular general Fintech applications of AI (e.g., customer service ‘chatbots’), what are potential significant applications specific to Private Securities? First, understand that ‘AI’ includes many different technical forms including from basic rule based systems that efficiently execute specified instructions, to machine learning systems that can discover new patterns from real time data streams.
Portfolio Analysis (Roboadvisors)
Especially for sophisticated investors it is important to consider not just individual investments but an overall portfolio strategy – stocks, bonds, alternatives, etc. what portion of each based on the investor’s goals. Increasingly the functionality of investment advisors (RIAs) is being incorporated into AI systems that provide analysis, either directly to the investors, or used by RIAs for their clients. Roboadvisors are digital platforms that provide automated, algorithm-driven financial planning services with little to no human supervision. A typical roboadvisor collects information from clients about their financial situation and future goals through an online survey, and then uses the data to offer advice and/or automatically invest client assets. This technology can reduce the cost of delivering portfolio advice to larger audiences of investors and is especially attractive to technically inclined investors interested, for example, in digital securities and peer-to-peer lending. This can be achieved with relatively simple rule-based AI that just applies standard portfolio analysis with greater efficiency,
Valuation and Risk Assessment
As mentioned, the theory of private placements is that they are restricted to sophisticated investors as they are able to tolerate greater potential risk than the general public, but this does not mean they are not risk sensitive. AI can apply not just at the initial investment valuation stage, but also may make possible continual monitoring of counterparty credit risk and operational risk for the issuer entities. This would involve more sophisticated AI techniques that enable real time pattern analysis, to perform complicated analysis.
Fraud Detection
With the increasing movement to digital securities, complicated offerings, automated trading and high-velocity transactions, there is an increasing development of AI technology to enable fraud detection. This is the subject of considerable research with the regulatory agencies. For example, the SEC is using both ‘supervised learning’ machine learning to identify potential fraud patterns in the text of SEC filings, and it is reported that these techniques are five times better than random in finding language that merits enforcement referral. Unsupervised learning algorithms are used by the SEC to identify unique or outlier reporting behaviours.
Conclusion
Private Placement investments offer sophisticated investors the ability to invest in equity, fractional loans and derivatives, with a potential for attractive returns, but these investments are restricted to accredited investors and qualified purchasers because they are exempt from normal securities filings and may present additional risks. Understanding these risks and having a clear understanding of the underlying investments value is critically dependent on the full ecosystem of financial securities professionals – broker/dealers, compliance, legal, exchanges, custodians, etc., and these are being extended and augmented by the advent of digital securities.
In the same way that digital securities can offer additional transactional efficiencies over traditional offerings, artificial intelligence mechanisms – from simple rule based advisory systems to real-time machine learning risk analysis and fraud detection – also offer significant new fintech implementations and improvements on historical manual processes. A corollary to the indelible first principle of focusing on the value of the investment is: “Make sure the counterparties to the investment are as good as the investment itself”.
