Marking-to-Model

Marking-to-Model (also known as 'mark-to-model') is a valuation technique used to estimate the fair value of an asset when a ready market price is unavailable. It’s the educated guess that accountants and fund managers have to make when they can't simply look up a price on a screen. This method stands in direct contrast to its more straightforward cousin, marking-to-market, where assets are valued at their current, observable market price. Marking-to-model is typically reserved for complex, illiquid, or unique assets that don’t trade frequently, such as certain types of bonds, exotic derivatives, private equity holdings, or even venture capital investments. The process involves creating a financial model that uses a range of inputs and assumptions to arrive at a theoretical value. While necessary for a functioning financial system, this method is famously controversial because its outputs are only as reliable as the assumptions plugged in, a fact that played a starring role in the 2008 financial crisis.

Imagine you own a one-of-a-kind, hand-carved sculpture. How much is it worth? There’s no daily market for it, so you can't just check the latest price. You'd have to estimate its value based on factors like the artist's reputation, the cost of materials, and prices of somewhat similar sculptures. Marking-to-model is the financial world’s version of this. Instead of a sculpture, think of a complex financial instrument. To value it, an analyst builds a mathematical model. This model is like a recipe, and it requires specific ingredients (inputs) to work. These inputs can include:

• Current and projected interest rates.
• The volatility of a related asset (e.g., a stock or a commodity).
• Projections of future cash flow an asset might generate.
• The creditworthiness of the parties involved in the transaction.

A common technique used in marking-to-model is a discounted cash flow (DCF) analysis, where analysts project an asset's future earnings and then discount them back to what they’re worth today. The crucial takeaway is that the final valuation is not a hard fact but an opinion—a highly educated one, perhaps, but still an opinion shaped by the model's design and the assumptions of its creator. As the old computer science saying goes: “garbage in, garbage out.”

Marking-to-model is a double-edged sword. It can be a useful tool or a weapon of financial self-destruction, depending on how it's wielded.

In a world of financial innovation, not everything fits into a neat, publicly traded box. Without marking-to-model, it would be impossible to assign a value to many assets on a company's balance sheet. This would stifle innovation and prevent companies from investing in unique opportunities. It provides a consistent, albeit theoretical, framework for valuing the un-valuable, allowing markets to function for assets that would otherwise be too opaque to handle.

Herein lies the greatest risk. The models are run by humans, and humans can be biased, overly optimistic, or simply wrong. By tweaking a few key assumptions—projecting slightly higher growth, assuming slightly lower risk—an analyst can dramatically inflate an asset’s value. This can lead to a distorted picture of a company's health, generating phantom profits and undeserved bonuses for managers. Because of its subjective nature, two different banks could value the exact same asset at wildly different prices, turning financial statements into a work of fiction rather than a statement of fact.

The 2008 financial crisis is the ultimate cautionary tale for marking-to-model. Banks were sitting on mountains of complex mortgage-backed securities, like collateralized debt obligations (CDOs). When the U.S. housing market seized up, the market for these assets vanished overnight. There were no buyers, so there were no prices. Forced to value these toxic assets, banks turned to their internal models. Unsurprisingly, these models were often built on wildly optimistic assumptions that housing prices would never fall on a national scale. This allowed institutions like Lehman Brothers and others to hide the terrifying scale of their losses, keeping asset values artificially high on their books. When the models finally broke and the true, catastrophic losses were revealed, the global financial system was pushed to the brink of collapse.

For a value investor, the concept of marking-to-model should set off alarm bells. The philosophy of value investing is built on a foundation of certainty and prudence, best embodied by the principle of margin of safety. You buy an asset only when you are confident of its intrinsic value and can do so at a significant discount. Valuations derived from complex, opaque models are the polar opposite of this. They introduce a huge element of uncertainty and subjectivity. The legendary investor Warren Buffett famously warned, “Beware of geeks bearing formulas.” He was cautioning against precisely this: relying on complex mathematical models that can obscure risk and create a false sense of precision. When analyzing a company, pay close attention to the portion of its assets labelled as Level 3 assets in its financial statements. This is the official accounting category for assets valued using marking-to-model. A large proportion of Level 3 assets is a red flag. It signals a high degree of complexity and estimation, making the company's reported earnings and book value less reliable. For the prudent investor, simplicity and transparency are king; a business whose value can only be guessed at by a computer model is often a business to avoid.