All posts
The S&P 500 has closed at a record 27 times so far in 2026. The latest came on 13 August, and in late September the index was still just over 1% below it.
Every one of those highs brings the same worry about investing at all-time highs back into the conversation.
Then something happens to otherwise sensible people.
A bonus, an end-of-service gratuity or the proceeds of a flat sale in London was heading for a portfolio. Instead, it stays in a deposit account for another month or two, until things 'settle down'.
The reasoning feels sound. Prices have climbed a long way, so there seems to be more room to fall than to rise. Why not wait for a dip and buy cheaper?
I understand the instinct. Wanting a better price feels perfectly reasonable.
It rests on a misreading of what a record actually is, though, and it's tended to cost people more than the fall they waited for would've saved.
So should you invest at an all-time high, or wait for the market to fall first?
If the money is surplus and your plan says it belongs in the market, a record high is no reason to wait. The principle we work to at AES is called investing by sunset. Money goes to work as soon as it's available, rather than waiting for a better day.
I made a shorter version of this argument in my column for The National. This is the fuller case, with the evidence behind it.
Before the detail, the points that matter most:
- If your plan says the money belongs in the market, a record high is no reason to hold it back. Dimensional found S&P 500 returns after a new high were close to the average after any month from 1926 to 2022.
- Records are routine in a market that rises over time. Around 30% of the S&P 500's month-end closes between 1926 and 2022 were new highs.
- Waiting has usually been the expensive option. Vanguard studied the MSCI World Index from 1976 to 2022 and found a lump sum invested at once beat a year in cash 70% of the time.
- Investing by sunset settles the timing question. That frees your attention for the decision that matters far more: which money belongs in the market at all.
What a record high tells you about the market
How often do stock markets hit record highs?
Stock markets hit record highs far more often than the word suggests. In Dimensional's study of more than 1,000 monthly closes for the S&P 500 from 1926 to 2022, 30% were new highs.
In sport, a record is rare, and that's why it's celebrated. A world record can stand for decades.
A market that grows over long periods behaves differently. If the direction of travel across decades is upward, the market spends much of its time passing its old high and setting new ones.
Records also tend to arrive in clusters, sometimes for months on end. Then they can disappear for years. Neither pattern tells you much about what comes next.
Does a record high mean a fall is coming?
A record high on its own says very little about whether a fall is coming. It describes where prices are today, and the next move could go either way.
A decline can begin at any time, and most begin unexpectedly, which is rather the point of them.
Dimensional's figures show the S&P 500 was higher a year after a record 81% of the time. Five years on, it was higher 86% of the time. After a new high, the average annualised return was almost 14% over the next year, and a little over 10% a year over five years.
Those returns were close to the average after any month, record or not. They're US figures in US dollars, and past performance is no guarantee of future returns.
The cost of waiting: investing at all-time highs versus holding cash
Waiting can feel prudent. For many investors it's been expensive, and the expense is easy to miss because nothing visibly goes wrong.
What does waiting for a dip actually cost?
Waiting for a dip costs you whatever the market returns while your money sits in a deposit account. The long-run return from shares is paid only to the people who own them, on the days they own them.
The fall you're waiting for might not arrive for a long time. When it does, it may start from a level well above today's price.
Vanguard measured the cost in a February 2023 research paper on cost averaging. Using MSCI World Index returns from 1976 to 2022, it found a lump sum invested at once beat a year in cash 70% of the time.
In roughly seven rolling one-year periods out of ten, the investor who held back finished the year behind. The growth happened somewhere they weren't.
The bargain rarely feels like one when it arrives
If a fall does come, it'll arrive wrapped in bad news. Bad news is usually what makes prices fall in the first place.
So the buying opportunity you pictured rarely feels like one when you're standing in it. The headlines are grim, and the fall could go further.
At that point the decline itself starts to look like a reason to wait for a bigger one. Once you're inside that loop, investing with any confidence becomes very hard.
The best investors I've known accept that their timing will never be perfect, and invest anyway. They embrace the uncertainty that systematic investing rewards, rather than trying to outguess it.
Is it better to invest a lump sum or phase it in?
Investing a lump sum at once has usually done better than phasing it in, but phasing in on fixed dates has still beaten holding cash. Vanguard found a lump sum beat three equal monthly instalments 68% of the time, and the instalments beat cash 69% of the time.
Phasing in on fixed dates is known as pound cost averaging, or dollar cost averaging in the US. The longer the phasing period, the higher the cost, because more of the money sits in cash for longer.
Here's how the three approaches compare, using Vanguard's figures:
| Invest by sunset (lump sum) |
Phase in on fixed dates | Hold cash and wait for a fall | |
|---|---|---|---|
| When the money is invested | As soon as it's available | In equal parts over a set period, often three months | When you judge prices have fallen far enough |
| What decides the timing | Availability of the money | A schedule fixed in advance | Your read of the market and the news |
| Historical record (MSCI World, 1976 to 2022) | Beat three-month phasing 68% of the time; beat cash 70% | Beat cash 69% of the time | Behind both approaches about seven times in ten over one year |
| What can go wrong | Fully invested just before a fall | Part of the money still in cash as prices rise | The fall never comes, or starts from a higher level |
| Who it can suit | Surplus money your plan says belongs in the market | Investors who would otherwise not invest at all | Money you'll need soon, which shouldn't be in shares anyway |
Vanguard's figures assume a 100% equity portfolio, a one-year horizon and no interest on uninvested cash. Capital is at risk, and past performance is no guarantee of future returns.
I go into this in more detail, with the charts, in my piece on whether to invest a lump sum all at once or gradually.
Investing by sunset: the principle and how to apply it
What does 'investing by sunset' mean?
Investing by sunset is the AES principle for surplus cash: money goes to work as soon as it's available, rather than waiting for a better day. It takes market timing out of the decision altogether.
The principle draws on decades of research into how markets price information. Eugene Fama shared the 2013 Nobel prize in economics for that work.
Fama showed that short-term stock price movements can't be predicted, because new information affects prices almost immediately. A record is information the market already has.
The name is a figure of speech, not a deadline. It simply means investing the money when you have it, rather than waiting for a better moment.
It applies to money your plan says belongs in the market. It doesn't apply to money you'll need soon.
When investing everything at once feels like too much
If handing over a large sum in one go still makes you uneasy, there's a middle path. You commit now to investing it in equal amounts on set dates over the coming months. What matters is the dates are fixed in advance, and not left to how the market feels on the day.
Vanguard's research suggests keeping that period short, around three months, because the cost grows the longer money waits.
Feeling uneasy at a high is completely normal. Investing well runs against instincts that evolution gave us for very different problems.
Why the timing is the smallest decision you'll make
More important than any market forecast is the part within your reach: deciding which money belongs in the market at all.
Money you won't touch for years belongs somewhere quite different from money you'll need soon. A house deposit due next spring has no business in shares, at a record or anywhere else.
Getting that separation right does more for your long-term result than any attempt to guess the market's next turn. It's the least interesting decision in investing, and comfortably the most valuable.
The separation can't be made from a chart. It comes from knowing what the money is for, when you'll need it and what rate of return your plan actually requires.
That's the order AES works in, and it sits at the centre of Financial Life Management: purpose first, then the plan, then the portfolio. Once the plan says which money is surplus, investing by sunset decides the timing, and the market's latest record stops being a question.
Once you know what the money is for, the timing mostly takes care of itself.
Where this leaves you
Nobody knows what the market will do next month. I don't, and neither does anyone offering a confident forecast.
What the evidence does say is a record high has told investors very little about what comes next. Waiting in cash has usually cost more than it saved.
There's no guarantee in any given period. Over long periods, though, markets have rewarded people who own a diversified mix of shares and stay owners.
That reward is measured in future purchasing power, which is what pays for the things that matter to you.
So if the money is surplus and your plan calls for it, a record is no reason to wait. If you'll need the money soon, it doesn't belong in shares at any price.
Common questions about investing at all-time highs
Should I sell my investments when the market hits a record high?
Not because of the record alone. A new high has said little about what happens next. Selling also leaves you deciding when to buy back in, which is the same timing problem twice. Rebalancing back to your target mix after a strong run is a different matter, and a sensible one.
Is holding cash at today's interest rates a good reason to wait?
Usually not on its own. Vanguard found that even allowing for interest on cash, investing at once beat a three-month phase-in 65% of the time. Higher cash rates do narrow the gap. Cash you'll need soon belongs in cash whatever the market does.
Does a record high mean shares are overvalued?
Not by itself. A record describes the price, not the price relative to company earnings. I covered overvalued markets in this video.
How do I know which money belongs in the market?
It depends on when you'll need it. Money for the next few years, including a cash reserve for emergencies, generally belongs in cash or something similar. Money you won't touch for many years is what investing by sunset applies to. Where that line sits is a question for your plan.
If you'd like to talk through what to do with cash that's waiting for a better day, book a 15-minute Discovery Call.
