Morgan Stanley Slip Up? Here’s Why Analysis Should Be Automated

Last week Morgan Stanley published an equity research note on Snap, the company it helped take public, putting a $28 price target on the stock, only to correct it a few hours later. But surprisingly, the correction, which included some changes in important metrics such as EBITDA and WACC, did not change the target price and the recommendation that followed: Buy a SNAP stock.

The story was picked and broadly covered by the New York desk of Business Insider, describing in detail what might have happened. After admitting the calculation error, Morgan Stanley cut Snap’s 2025 adjusted EBITDA from $6.57 billion to $4.92 billion, and the free cash flow from $4.05 billion to $2.42 billion. That’s a drop of $1.7 billion in EBITDA and $1.6 billion in free cash flow. As the classic model of valuing a company demands using discounted cash flow, growth, and profit in estimating a company’s valuation, you would expect Morgan Stanley to change Snap’s stock price target. But it remained the same.

After maintaining its recommendation to buy Snap’s stock last Monday, shares rose 4% in New York. Morgan Stanley’s ‘buy’ rating was widely covered and was also mentioned as one of the reasons, together with four other analysts, for that day’s hiccup in stock price. Morgan Stanley was not just another analyst – it was Snap’s lead underwriter.

Source: Yahoo Finance

How did Morgan Stanley maintain the same price target of $28 despite the fact that it was higher by 23% than the previous business day’s price? Business Insider explains that it did so by lowering Snap’s equity risk premium and the weighted average cost of capital (WACC), a measure that takes into account the cost of issuing equity and borrowing, a figure that is considered as highly subjective in a valuation model.

Was that a coincidence? Did Morgan Stanley find an inaccuracy in either EBITDA and WACC as well, that allowed to maintain the same price target?  It may look as if they were backing into the numbers, some experts are convinced, but it is actually not the real issue. The thing is that the main model of business research and equity research is broken. And yet, every brand new technology company that goes public is still dependent on old-fashioned conglomerates of investment, underwriting and research.

A New Research Is Needed

First and foremost the current research for public tech market is skewed. It is ruled by investment banks that are market makers, rooted deep in Wall Street (and the Silicon Valley). For many research companies, research is just a platform, a product that facilitates the upsell of a panoply of other services, such as consulting, conference and workshop management, and publication rights. And research is expensive, offered by banks or giant research firms for somewhere between $30K on “access for a research platform” up to $10M “to provide a manager access to across the research company’s workforce.”

One of the reasons research is so expensive today is that it employs researchers who do their job using century-old, outdated methods and mindsets. And, yes, people sometimes can get things wrong and – worse – sometimes they play with the numbers. We don’t claim that is the case in Snap’s March 27th metrics correction, we have no idea what happened there, but there are enough other cases to prove this point.

It must not be that way. As AI and machine learning technology evolve, developing greater capabilities and replacing human labor in manufacturing, advertising, e-commerce, and even healthcare, these new technologies come to play a more prominent role in financial services. As algorithms help bankers hedge risks by better profiling loans or mortgage consumers, and as bots contribute to providing faster, more personal service, technology can assist in making equity research better.

That’s what we’re doing here at Zirra. We provide insights on private companies, which are synthesized from metrics produced by our proprietary AI and machine learning algorithms, as well as insights garnered from our network of experts.

Companies’ valuation, for instance, is calculated by a myriad of algorithms that automate the traditional valuation method. The process includes both Intrinsic and Relative valuation algorithms. The Intrinsic data includes revenue and expense estimations, traffic trajectories, advertising campaigns measurements, investment history, and velocity, based on aggregated sources. In the Relative analysis, data is compared and benchmarked with a database of thousands of companies with correlation to stage, space, size and trajectory. In total, we use more than 85 different public data sources to feed our system.

Search for a company

This also produces a preliminary set of company ratings for other aspects of the company such as the product, the momentum behind the company and the founding team. For example, the algorithm concluded that teams of up to 3 co-founders, each specializing in his/her field of business, technology or marketing leads to growth, whereas teams of 4 or more co-founders carry a higher degree of risk. The machine learning algorithms track the growth of companies, arrive at conclusions independently and dynamically adapt for future analysis.

The proprietary big data technology allows us to scale and evaluate startups metrics within seconds and our network of experts allows us to create qualitative and valuable insights that no other directory can create. It’s faster; it’s tailor made to any startup or private company. And it’s much cheaper.

Investment banks are also starting to adopt machine learning solutions for various purposes. JPMorgan launched a predictive recommendation engine to identify those clients which should issue or sell equity. Also, Goldman Sachs’ execs said they were willing to develop an automated IPO process and other investment banking tasks.

Technology-based research is also more transparent and less prone to tricks as traditional research is. Oh, and Snap’s valuation by Zirra is $26 billion, as it has been for months prior to the company going public. This is also its current valuation today at the stock exchange. Try it here, our data base is free.

Assaf Gilad

An ex-journalist from Calcalist, a leading business and tech news outlet in Israel, I'm now writing about startups for