Ceci N'est Pas Custom

Why Custom Direct Indexing is not custom and how you can access true portfolio personalization

Introduction

The evolution of technology constantly reframes consumer expectations, thereby forcing RIAs to adapt. Fortunately, that same technology often gives way to tooling that enables advisors to meet the needs it creates. For example, when automation gave rise to robo-advisors it increased consumer expectations around ease of use and transparency. That same technology also produced rebalancers, which save advisors time and allow them to put greater emphasis on client experience to meet these new expectations.

The latest chapter of this storied saga is defined by personalization. Once again, consumer expectations have been reset as services such as Amazon, Spotify, and Netflix provide increasingly tailored offerings through machine learning and artificial intelligence. This has resulted in clients requesting personalized services from their financial advisors more than ever before. Deloitte refers to this trend as hyper-personalized wealth management¹, but we simply see it as the extension of technologically enabled individualism into the world of finance.

Clients now want services tailored to their unique needs, preferences, and financial goals, rather than one-size-fits-all solutions. Many advisors have turned to Custom Direct Indexing tools to meet these demands. Unfortunately, a lot is left to be desired. According to eMoney, 60% of consumers are dissatisfied with advice that is too generic for their situation².

The core problem is that the technology that has redefined consumer expectations is not the technology that is being used by RIAs to meet those new client demands. Custom Direct Indexing is built on an antiquated understanding of customization and portfolio management. To truly meet the needs of the modern client, RIAs must embrace tools that harness cutting edge machine learning and AI technology, or they will fall behind other industries and, therefore, lag behind consumer needs.

What Is Wrong with Custom Direct Indexing

Custom Direct Indexing is the latest instalment in portfolio management methodologies that promise more than they can deliver. It represents a common trend in the financial industry in which the same, antiquated technology is rebranded and called innovative. That is a bold claim and not one we posit lightly, so let us break it down.

There are two main requirements for a portfolio to be personalized:

1. It has to meet the client’s surface level requirements, such as which sectors, industries, and assets are included and excluded
2. It has to meet the client’s performance requirements in terms of risk and return

In short, the issue with Custom Direct Indexing is that it satisfies the first requirement but fails to meet the second.

Custom Direct Indexing works by breaking a market benchmark, such as the S&P 500, into its component assets and then offering advisors the ability to add or remove assets to create a “customized” version of that benchmark. The problem with that approach is that as the portfolio deviates from the benchmark, its underlying exposures start to change, which can have a material impact on the portfolio’s risk and return characteristics. As a result, advisors and clients receive portfolios that appear personalized but have not addressed the necessary underlying changes.

To properly customize a portfolio, a tool needs to enable advisors to make surface level adjustments without sacrificing the underlying portfolio characteristics. For example, when an advisor removes tobacco stocks that do not align with a client’s values, the other assets in the benchmark need to be readjusted and optimized to minimize tracking error. However, with a portfolio built through Custom Direct Indexing, this step is left out entirely.

Why does Custom Direct Indexing ignore this step? Portfolio optimization is highly complicated and requires an in-depth understanding of the client’s needs, the assets within the portfolio, and the current market conditions. Custom Direct Indexing is based on an improper understanding of the first of those three. Portfolios built this way may satisfy client needs in the short term as they appear to meet their desires, but are sure to disappoint as time goes on and performance deviates from expectations.

If you’ve read or seen the movie Moneyball, you may remember the scene where scouts for the Oakland A’s are sitting around a table trying to find replacements for star player Jason Giambi. They go back and forth discussing player characteristics, such as having a great face or an ugly girlfriend. Of course, these immediately seem like silly factors to evaluate players on, but baseball used them for decades and finance is still using that kind of logic today. Custom Direct Indexing is incapable of looking beneath surface level factors to build truly personalized portfolios. Similar to how Oakland A’s GM Billy Beane brought the MLB into the 21st century, wealth management needs new tools to usher in the next generation.

Why Pave Pro Is the Future

If Custom Direct Indexing is technologically disadvantaged at its core, rendering it incapable of truly satisfying client needs, is it possible to build a system that can? Fortunately, the answer is yes, but doing so is far from straightforward.

As mentioned above, portfolio optimization requires an in-depth understanding of three elements:

1. The client’s needs
2. The assets within the portfolio
3. The current market conditions

Let us start with the first element: the client’s needs. There are several dimensions over which a portfolio can be personalized including values, circumstances, and life goals. Values are predominately about including or excluding certain assets that a client may or may not want to support. Circumstances can mean including certain assets a client earns as part of their compensation package as well as their tax situation. Life goals ultimately come down to the risk and return profile that best matches the client’s risk tolerance, retirement timeline, and ideal level of wealth.

Our first challenge arises when considering that properly satisfying all of these requirements necessitates any system to ingest, understand, and act on a tremendous amount of data. This requires complex infrastructure, but for the sake of simplicity, let us assume this part is done. Now, we move onto the different layers of personalization and how to handle them. Solving for values based personalization is relatively straightforward, and in a vacuum, Custom Direct Indexing does a good job of this. All that is required here is ensuring that certain assets are included or excluded from the final portfolio. As a result, we can treat this problem as being solvable with simple constraints. The complicated part comes next when wanting to simultaneously solve for taxes, risk, and return.

To successfully build a portfolio that fits the client’s value based constraints (asset inclusions and exclusions) while considering their tax situation, fitting their risk tolerance, and maximizing their potential return, the system needs to draw from all three elements referenced above. First, it needs to know the tax implications of selling assets as well as any specific requirements around the client’s desired tax burden. Second, it needs to have a clear understanding of the client’s risk tolerance and how to translate that into a mathematical target. Finally, it needs an understanding of the underlying assets, their risk and return characteristics, and how the market is rewarding or penalizing certain characteristics at the given moment as well as how it will do so into the future. Once it can do all of that, it needs to solve for all three simultaneously to generate a truly personalized portfolio. It is also critical to have the ability to generate these calculations periodically to ensure the portfolio stays well positioned for current client and market circumstances in perpetuity.

Hopefully, now it is evident that Custom Direct Indexing is simply scratching the surface of what it means to “customize” a portfolio. There is a happy ending to this story though. Pave Pro has solved every single problem laid out above and more.

Pave Pro can connect directly into an RIA’s custodian, meaning anyone can get up and running in less than two weeks without moving any client assets. Advisors can then enter all of the aforementioned personalization parameters and with the touch of a button receive a bespoke portfolio for each one of their clients in a matter of seconds. Pave Pro also actively manages portfolios on an ongoing basis and is capable of automatically trading in and out of positions in a tax efficient manner without requiring the advisor to lift a finger.

By harnessing an asset scoring engine used at a leading bulge bracket private bank in concert with cutting edge machine learning technology, Pave Pro is able to offer true portfolio personalization. Pave Pro’s asset scoring engine powered a strategy that outperformed the S&P 500 by an average of over 3% per year for over a decade while managing billions of dollars.³ Now it is proprietary to Pave, scores every publicly traded asset in the world on a weekly basis, and gives us an unparalleled understanding of risk and return characteristics at the individual asset level. On top of our scoring engine, we layered a machine learning system that predicts market trends and, therefore, understands what characteristics are likely to perform well going forward. Finally, we built our own optimization engine that takes in all of that asset and market data along with a client’s preferences and asset base to generate a one-of-one portfolio suited to every dimension of a client’s desires.

To provide a more concrete understanding of how Pave Pro functions, let us use the example of Client A, a high-net-worth individual who bought several tech stocks early and now has a portfolio with a multi-million dollar unrealized gain. Client A would also prefer not to invest in tobacco and alcohol companies. Faced with these circumstances and set of preferences, Pave Pro would first identify the assets in the client’s current portfolio that do not meet their requirements or are poorly positioned given the client’s risk tolerance and current market conditions. After identifying these assets, Pave Pro would analyze how selling the assets would benefit portfolio performance and weigh that against the tax impact of realizing the gains. Tobacco and alcohol assets would be sold immediately. The tech stocks, however, will only be sold if the performance benefits outweigh the tax implications. Pave Pro would also attempt to offset realized gains by realizing losses if those assets do not offer an outsized performance benefit. The entire sell list would be checked for turnover as well to ensure only the highest impact trades are being made. Once the sell list is finalized, Pave Pro would identify new or existing assets to buy in order to maintain a portfolio that aims to maximize returns, control the client’s risk, and suit the client’s preferences. All of this takes place in a matter of seconds before the advisor is shown a list of suggested trades and can place them at the touch of a button.

In 2021, The CFA Institute published a view of what they think a financial advisor’s workflow will look like in 2038⁴. They envisioned a world without model portfolios where advisors could tailor strategies to their client’s unique needs. Unfortunately, this future will not be realized through Custom Direct Indexing. In fact, it has already been realized by Pave Pro. The future is 15 years early, and it’s brighter than anyone could have imagined.

Contact Us

If you’re interested in learning more about Pave Pro, please visit pavepro.ai or contact us directly at salessupport@pavefinance.com. We look forward to hearing from you.

About the Author

Pascal Cevaer is the Founder and President of Pave Finance, Inc. Prior to Pave, he advised several Fortune 500 financial institutions as a part of McKinsey’s New York office. Pascal also spent time in Silicon Valley leading product initiatives specifically around the implementation of machine learning and AI systems at large companies and startups. He graduated from Stanford University with a B.S. in Computer Science and Product Design.

¹ Disrupting Wealth Management – The Age of Hyper-Personalization
² What Do Clients Want from Their Financial Advisor?
³ See full performance disclosures here
⁴ Wealth Management in the Algocen Era: A Speculative Future

‍