Business development companies like Ares Capital Corporation (ARCC 0.24%) are actually pretty simple creatures. They essentially act as arbitragers, sourcing capital from their stockholders and creditors on Wall Street to make debt and equity investments in private businesses in the United States and around the world.

The business isn't fundamentally that different from the traditional banking industry; it's all about using capital at a low rate to lend and invest at a higher rate. Done correctly, the result can be tremendous for shareholders, as double-digit dividend yields are relatively common in the industry.

Photo of a spilling coin jar

Don't get used to Ares Capital's large dividend; a cut is coming. Image source: Getty Images.

But because BDCs are required to pay out virtually all of their income, dividends are always at risk of a cut. One of the best ways to see whether a BDC's dividend is sustainable is to estimate the returns it needs to generate on its portfolio to earn its dividend. Luckily, almost all of a BDC's expenses are knowable, or at least estimable. Therefore, building a high-level "model" for the company isn't all that difficult.

My inputs for Ares Capital are shown in the table below:

Inputs

Figure

Cost of equity 

9.2%

Cost of debt

4%

Management and incentive fees

1.5% of assets plus 20% of returns ($10 million of quarterly fees waived)

Other SG&A expenses

0.5% of assets

Leverage

0.75 debt to equity

Source: SEC filings and company presentations. Notes: Cost of equity is calculated by the dividend divided by net asset value, cost of debt is the weighted average stated rate on principal amounts borrowed. Other SG&A expenses and leverage were estimated based off the prior three years.

With these inputs in place, I calculate that Ares Capital Corporation would need to generate a return on its portfolio of approximately 10%, including any fees, gains, and loan losses, to continue paying its current dividend of $0.38 per share, per quarter.

Generating this kind of return will be difficult, if not impossible. Consider that the weighted average yield on its total investment portfolio tallied to just 8.2% at fair value at the end of the first quarter. A more favorable figure -- the yield of its income-producing investments only -- stood at just 9.4% last quarter, still less than the 10% return on assets it needs to maintain its current dividend.

To be fair, Ares does have some ways to improve its portfolio yield. Ares identified $3 billion of assets yielding less than 7% that can be redeployed into higher-yielding investments. But redeployments aren't enough on their own, as the company will need to expand its portfolio with more than $1 billion of additional investments in order to walk up its leverage back to 0.75 times its equity, up from a recent leverage ratio of 0.64 times. Ares Capital finds itself in the disadvantaged position of having a lot of money to deploy when credit spreads are near historical lows.

Yield compression

Data from S&P Global Intelligence suggests that yields on riskier second-lien loans fell to 10% in the second quarter of 2017, down from more than 11% last year. Across the board, credit spreads are tightening, increasing the potential that Ares' current borrowers refinance at a lower rate, negatively affecting the return it earns on its loan portfolio.

And while Ares has proven to be a capable underwriter over its corporate history, I'm not willing to give it much credit for its ability to generate below-average loan losses in the short-term. Given where we are in the cycle, it seems much more likely that losses exceed gains, and thus Ares Capital's return on its portfolio will likely be lower than the 9.4% yield it is currently earning on income-producing assets, and certainly less than the 10% return on assets it needs to maintain its current payouts to shareholders.

Barring a permanent fee reduction, or an immediate and sharp increase in credit spreads, Ares Capital's dividend is destined for a cut. A cut to $0.30 quarterly from a current level of $0.38 isn't out of the cards, as the longer it waits to reset the dividend, the deeper the cut will be.