Lessons from 4 Years of GSE Preferred Trading

On paper, holding 23 different preferred series across Fannie Mae and Freddie Mac looks diversified. It wasn't. Every single position was the same bet: that the government-sponsored enterprises would exit conservatorship, and preferred shareholders would be made whole or close to it. Owning FMCCS, FMCCJ, FNMAS, and FNMAT simultaneously didn't reduce risk — it concentrated it into a single binary outcome.

The clearest example is FMCCS. Over four years, I accumulated 143,154 shares and sold exactly 100. That's not a trading position — that's conviction expressed in share count. The buy prices ranged from $2.80 to $18.91, meaning I was adding at nearly every price level the market offered. When you believe in the thesis that strongly, the ticker symbol is almost irrelevant. You're buying the outcome.

The lesson: don't confuse ticker-level diversification with thesis-level diversification. If all your positions win or lose together, you have one bet — no matter how many line items are in your brokerage statement. Own that reality. Size accordingly.

Between April 2022 and March 2026, I was a net buyer in 14 of 16 quarters. That means for nearly four straight years, more cash went into GSE preferreds than came out. Some months I made 500+ trades; others I made zero. But the direction was always the same: accumulate.

Prices went from $1.88 (the lowest FNMAT print in December 2022) to over $22 when FMCCJ, FMCCN, and FMCCM hit their highs in late 2025. That's an 11x move from trough to peak. But getting there meant holding through years where nothing happened — no catalyst, no legislative movement, no dividend, no news. Just silence and the occasional 20% drawdown to test your resolve.

The math on patience is asymmetric. If you sell during a drawdown, you crystallize a loss and need to find another idea. If you hold and the thesis plays out, the drawdown becomes invisible on the chart. The hard part isn't the analysis — it's sitting still while everyone around you trades momentum and you hold a security that hasn't moved in 18 months.

In the summer of 2022, I took a momentum position in QualTek International (QTEK) — 315,873 shares at an average price of $1.35. This was the one non-GSE stock I traded in any real size. It was a quick, tactical bet: buy the dip, ride the bounce, get out.

I sold everything by October 7, 2022, at an average price of $1.66 — roughly a 23% gain in 74 days. Four months later, QualTek filed for Chapter 11 bankruptcy. If I'd held because the position was "working" or because I'd developed an emotional attachment, those shares would have gone to zero.

The discipline here was simple: QTEK was never the thesis. It was a side bet, and side bets get closed when they've served their purpose. The GSE preferreds were the thesis — those I held for years. Knowing which positions deserve patience and which deserve a quick exit is maybe the most underrated skill in investing.

My closed options record is 1 win and 8 losses. An 11% hit rate. Eight positions expired completely worthless — MBI, GEO (four separate strikes), LUMN July calls, AMBC, and LDI. That's real money that evaporated on expiration day.

But the one winner — LUMN January 2026 $10 calls, 100 contracts bought at $0.40 — returned +293%. That single trade more than covered the cost of all eight losers combined. This is the nature of options: you're buying lottery tickets with asymmetric payoffs. The expected value can be positive even with a terrible win rate, as long as the wins are large enough.

The lesson isn't "don't trade options." It's that options only make sense when you have a high-conviction, time-bound catalyst and you're willing to lose 100% of the premium. They're a terrible tool for expressing vague optimism. Every GEO call I bought was essentially saying "this stock will rally 50%+ by this specific date." When it didn't, the premium was gone. Use options for asymmetric bets, not for leveraged directional trades.

A casual observer looking at my transaction log might see 2,068 trades and think "day trader." But the reality is different. The vast majority of those trades were rotations within the GSE preferred universe — selling FNMAS to buy FMCKJ, selling FNMAT to buy FMCCS, selling FNMAJ to buy FMCCP. The thesis never changed. Only the vehicle did.

Why rotate? Because different GSE preferred series trade at different discounts to par, carry different coupon rates, and respond differently to news flow. When FMCKJ was cheap relative to FNMAS, I'd swap. When fixed-rate series looked more attractive than variable-rate, I'd adjust. When a series had run up and another was lagging, I'd rotate capital. This is portfolio optimization, not speculation.

Over 100,000 shares of both FMCKJ and FNMAS were bought and sold entirely — every single share turned over. But the capital never left the GSE preferred thesis. Think of it like rebalancing between Vanguard funds vs switching from stocks to crypto. Same account activity, completely different intent.

The longest-hold positions were in tax-advantaged accounts. Roth IRA conversions from a SEP IRA were immediately deployed into GSE preferreds. This wasn't accidental — it was strategic. When your thesis has an uncertain timeline measured in years (not months), the last thing you want is a tax bill on unrealized gains forcing you to sell or complicating your holding psychology.

My Roth IRA accounted for 870 of the 2,068 total trades — more than any other account. The main individual account had 838. But the Roth held the positions that were bought earliest and cheapest, because there was zero tax incentive to trim. In a taxable account, a 5x gain tempts you to sell and lock in long-term capital gains. In a Roth, a 5x gain just sits there, compounding, with no tax consequence.

If you have a multi-year thesis — especially one where the capital gains will be substantial if you're right — the Roth is the best vehicle. You pay the tax upfront on the conversion (at the lower basis) and let the entire gain compound tax-free. My SEP IRA contributions went in, got converted, and got deployed within days. That's how you play a long game.

Here's something most investors can't stomach: holding a security that pays zero income for years. Every GSE preferred stock has had its dividend suspended since the 2008 conservatorship. That's 18 years (and counting) of no cash flow to shareholders. No quarterly check to rationalize the hold. No yield-on-cost to show off. Just a stock price, a thesis, and your conviction.

This is a powerful filter. It eliminates every investor who needs income, every fund with a yield mandate, and every advisor who can't explain to a client why they own a non-income-producing fixed-income security. What's left? True believers in the privatization thesis. And because the holder base is self-selected for conviction, the sellers tend to be weak hands — forced liquidations, margin calls, or people who simply lose patience.

Buying into zero-dividend securities requires a different mental model. You're not an income investor — you're an event-driven investor. The "dividend" is the capital gain when the event occurs. Every share I bought was priced as if the dividends might never resume. That's the discount I was buying. The lesson: the absence of income isn't a bug — it's the source of the opportunity.

FMCCN went from $2.27 to $22.00. FMCCJ went from $2.07 to $22.14. FMCCM went from $2.36 to $22.00. These are roughly 10x returns. They didn't require a complex strategy, sophisticated timing, or options leverage. They required buying at distressed prices and waiting for the thesis to play out.

The math on multi-baggers is simple but unintuitive. A stock that goes from $2 to $20 is a 10x. A stock that goes from $20 to $40 is a 2x. The same $18 of absolute appreciation, but the return profile is completely different. This is why distressed and special-situation investing produces outsized returns — the starting price is depressed by uncertainty, not by fundamentals. When the uncertainty resolves, the rerating is violent.

But here's the part nobody tells you: a 10x only works if you had meaningful size at the bottom. My average buy prices weren't $2.27 — they were higher, because I kept buying as prices rose. FMCCN's average buy was $3.88, FMCCJ's was $3.81, FMCCM's was $3.91. Still multi-baggers from average cost, but not 10x. The lesson: your entry price is everything in distressed investing, and your willingness to buy when it feels worst determines your ultimate return.