Fannie Mae and Freddie Mac guarantee most American mortgages, which means they eat the losses when homeowners default. Rather than carry all of that risk themselves, they quietly sell a chunk of it to private investors. Those investors collect a steady payment, and in exchange they agree to lose money when enough borrowers stop paying.

The instrument is called a credit risk transfer, or CRT. One of these programs alone has placed tens of billions of dollars of mortgage-default risk into the hands of bond funds, insurers, and hedge funds. The reader may own a slice of it inside a credit fund without knowing it.

The machine we are drawing is that transfer. It is a way to move the danger of a wave of mortgage defaults off the taxpayer and onto willing investors, and it works beautifully until the thing being insured, the American homeowner, starts to miss payments all at once.

The two companies have run this program since 2013, built after the 2008 crisis to stop taxpayers from being the only backstop for mortgage losses. Across its life it has transferred well over a hundred billion dollars of credit risk to the private market.

The scale is steady and large. At one of the firms, the maximum losses that CRT investors could be forced to absorb stood near $37 billion in mid-2026, a cushion standing in front of the taxpayer on a slice of the mortgages it guarantees.

Why this reaches the reader: this program decides how much of the risk on his own mortgage is carried by private investors versus by the public. And with the companies possibly heading out of government control, whether they keep sharing that risk, or quietly take it back, is suddenly an open question.

You won't hear this in the media…

But we are at the inflection point right before a new type of AI called "Accelerated AI" explodes into the mainstream…

And unlocks an entirely new dimension of exponential growth.

If history is any guide, we could be looking at potential gains of up to 10,000% from here.

If you want to find out more about "Accelerated AI" and get the name and ticker of our #1 stock pick — for free…

Click here now.

Start with the companies' exposure. Fannie and Freddie promise bond investors they will be paid even if homeowners default, so a surge in defaults lands squarely on the two firms, and in a crisis, on the taxpayers behind them. That concentration of risk in two companies is exactly what blew up in 2008.

The fix was to spread the risk. The firms take a large pool of the mortgages they guarantee, call it a reference pool, and sell bonds whose fate is tied to how those specific loans perform. Investors buy the bonds at full price and earn a healthy yield, funded by a slice of the fee homeowners pay for the guarantee.

The load-bearing part is the first-loss design. The bonds are stacked in layers, and the lowest layer absorbs the earliest losses. As homeowners in the reference pool default, the investors' principal is written down, bottom layer first. They are, in plain terms, selling insurance on strangers' mortgages and getting paid a premium to stand in front of the taxpayer.

Note what the investor does not get. He never owns the mortgages. He holds a bond whose value is linked to them, and whose principal shrinks as they sour. It is a synthetic bet on homeowners paying their loans, structured so the company offloads the danger while keeping the loans themselves.

This is why it can strain a giant. In calm times, defaults are low, investors collect their premium, and the taxpayer's exposure is genuinely reduced. But the risk has not vanished; it has been parked with investors who can flee. If they stop showing up to buy new deals, the firms are thrown back onto their own balance sheets, right when losses are rising.

The first joint is correlated default. These bonds are safe as long as mortgage losses stay scattered and modest. In a broad housing downturn, defaults cluster, losses climb through the layers, and the lower tranches can be wiped out. The investor is paid for ordinary risk but exposed to the rare, system-wide event, which is precisely when it hurts most.

The second joint is the buyer who walks away. The whole machine depends on private investors willing to buy new risk-transfer deals year after year. When fear spikes, as it did in early 2020 when pandemic defaults loomed, the price of this risk blew out and issuance stalled. If no one will buy the risk, it flows straight back to the companies and the taxpayer.

The third joint is the exit from government control. If Fannie and Freddie are released to private shareholders, they may choose to keep more of the risk, and the fat guarantee fees that come with it, rather than share them with CRT investors. That would hand profits to shareholders while quietly re-concentrating mortgage risk in the two firms, undoing the very diversification the program was built to create.

What persists: the diversification, which is real and valuable. Spreading mortgage losses across hundreds of private investors is genuinely safer than piling it all on two companies, and the program has performed well through a strong housing decade, paying investors and shielding taxpayers as designed.

What can break: the transfer, exactly when it is needed. The risk is only moved for as long as investors stay willing to hold it, and their willingness evaporates in a true housing crisis, the one scenario the program exists to cushion. The protection is strongest in good times and thinnest in the bad, which is the oldest flaw in any insurance that can be canceled by the insurer walking off.

This machine connects to two others the reader's money touches. It sits on top of the same Fannie and Freddie guarantee we have traced, deciding how much of that guarantee's risk is shared out, and it is a cousin of the catastrophe bond, another security that pays a premium until a disaster, here a wave of defaults instead of a storm, wipes the principal out.

The next reading arrives with each quarter's risk-transfer disclosures and every step of the conservatorship-release debate. Those will show whether the companies keep handing mortgage risk to private investors, or whether taking the giants public quietly loads that risk back onto the public that stands behind them.