Mathematical Money | October 2, 2026
The question people love to pose is a choice between two lives. Option one is around 5% a year, safe, limited drawdowns, sleep at night. Option two is 20%-plus a year, but you have to survive drawdowns of 50%. It's a good question and I was going to write it up as a straight either-or, until I went to check whether the first option actually exists. I don't think it does.
What I measured
I pulled daily closes for the usual "safe income" names and the usual "risky" ones over the same window — 20 December 2021 to 1 October 2026, which is 4.78 years and takes in a rate shock, a bear market and a full recovery. These are price returns and price drawdowns, so they exclude dividends, and I'll deal with that properly in a moment because it genuinely matters.
Start with the income side. Realty Income drew down 38.3% at its worst and has compounded at −4.7% a year on price, still sitting 28.6% below its peak today. The broad REIT ETF is almost identical at 38.8% and −4.5%, still 23.1% below. Utilities drew down 28.1% to deliver under 3% a year, and dividend equities were the mildest of the group — 18.9% worst drawdown for +4.9% a year, which is the one genuinely sensible result on this list.
Now the ones everybody calls risky. The S&P 500 fell 25.4% at its worst, compounded at 11.5%, and currently sits 1.8% below its high. The Nasdaq 100 fell 35.5%, compounded at 15.0%, and is 0.7% below its high.
Then there's long-dated Treasuries, which deserve their own sentence. Down 48.1% from peak, compounding at −12.8% a year, and that worst-ever drawdown is happening right now — the asset has never been lower in this entire window.
Read those numbers again
Realty Income is the archetype of the safe monthly dividend, yielding somewhere around 5.7%, and it is exactly what people picture when they say they'll take 5% and sleep well. Its maximum drawdown over the past five years was larger than the Nasdaq's. Not similar — larger. The income asset bought for safety fell 38.3% while the growth index everyone calls dangerous fell 35.5%, and the dangerous one compounded at 15% a year while the safe one went backwards.
That isn't a story about one badly run REIT either. The broad ETF did the same thing, utilities drew down 28% to earn you under 3%, and the only member of the defensive camp that behaved as advertised was the dividend-equity fund.
Being fair about dividends
Price-only comparisons flatter growth stocks and I don't want to cheat, so here's the adjustment. Realty Income has yielded roughly 5 to 6% through this period, which adds something like 25 to 28 percentage points over the full stretch and turns that −4.7% a year into roughly break-even, maybe a shade positive. The REIT ETF lands slightly negative. Dividend equities end up near 8.4% a year, which is genuinely respectable. The S&P and Nasdaq barely move because their yields are small.
So the honest version is that income assets didn't destroy your capital — they returned roughly nothing to modestly positive over nearly five years, while putting you through a drawdown of a third. That's still not the deal that was advertised. "Safe 5%" was sold as low return for low risk, and what arrived was low return with high risk.
It's also worth saying that the drawdown you actually lived through was the price drawdown. Dividends landing quarterly don't stop the number on your screen reading minus thirty-eight percent, and the psychological experience is the price chart, not the total return chart you construct afterwards to feel better.
It gets worse the further back you look
In 2008, US REITs fell somewhere around 70% peak to trough while the S&P 500 fell 57%. The asset class marketed on stability fell meaningfully harder than the stock market, in the one event everybody still uses as their benchmark for disaster. Closer to home, Singapore REITs dropped roughly 75% in that same crisis, and a lot of people here hold S-REITs precisely because a 5 to 6% yield feels like the conservative choice. On this evidence it isn't obvious to me that it is.
And then there's the one that should stop everybody. Long-dated Treasuries are the asset the entire industry calls risk-free, and they're down 48.1% from peak and sitting at the bottom of that range as I write. The safest thing in finance produced the worst drawdown on my list.
What "safe" actually means
Low volatility, until it isn't. Most things marketed as safe genuinely do move very little most of the time, and that's a real property worth something — but it tends to be bought with leverage, duration, or both, and those are exactly the characteristics that produce violent repricing when rates move. REITs are levered real estate, long bonds are pure duration, utilities are capital-intensive and rate-sensitive. All three got hit by the same force at the same time, which is also why owning all three diversified nobody.
The yield was never a gift. It was payment for a risk that simply hadn't arrived yet, and when it arrived it took more than five years of income with it in a few months.
So I'd reframe the whole question
The choice isn't between 5% with small drawdowns and 20% with big ones, because over the last five years essentially everything drew down somewhere between 19% and 48%. The drawdown wasn't optional in any of them. What differed was whether you knew the number before you bought.
Hold Realty Income expecting a worst case of maybe 15%, and 38% breaks something — usually your conviction, usually near the bottom, usually permanently. Hold the Nasdaq knowing full well it does this every few years, and 35% is uncomfortable but survivable, and you're back at the highs eighteen months later. Same drawdown, completely different outcome, and the variable was the expectation rather than the asset.
That's why I'd rather own something volatile whose behaviour I've measured than something calm whose behaviour I've assumed. The second is where people get genuinely hurt, because nobody plans for the drawdown they were promised wouldn't come.
So here's the question in the title, with the numbers attached. Both options cost you roughly a third at the worst point. One of them compounded at 15% a year and was back at its highs within eighteen months. The other paid you 5% a year in dividends, returned about nothing in total, and is still 28% below its peak five years later. Which one were you actually choosing when you picked the safe option?
And for anyone holding income assets right now, a more useful question than mine: do you know your position's worst historical drawdown, or do you only know its yield? Most of us can quote the second instantly and have never once looked up the first.
Stop guessing. Start calculating.
Live to fight another day. 🤙
Comments