Who says equality wasn’t a thing 30 years ago? When I first began as a portfolio manager my colleagues were evenly split between Cambridge and Oxford graduates. Can’t be fairer than that.

And our trading desk more than ticked the diversity box. An eccentric mix of characters and backgrounds it was — sitting smack in the middle of the office. When not at lunch, orders were shouted and phones slammed.

What did they do? And who cares anyway, you might ask. Well, they were crucial to our fund performance. Likewise, your retirement portfolio is reliant on traders to this day — even if more is automated now.

I will explain how with an example. When I was running Japanese equity portfolios, let’s suppose I wanted to purchase 100,000 shares of Nintendo. I would input the order into our system. It would then pop up on a screen in front of a bloke called Gara.

His job was to try to secure me the lowest price possible. And the market was in effect made by brokers — Nomura or Goldman Sachs or Deutsche Bank — who matched buyers with sellers.

First, Gara would check whether 100,000 was a large proportion of the average number of Nintendo shares traded per day. If it was, he might spread my order over a number of days so as not to spook the market and drive prices higher.

Alternatively, he could use more than one broker to hide our true size. It was also up to him whether to direct the broker in some way — much like we retail investors can do on our platforms. Buy at the open. Don’t go above such and such a price.

Or Gara could simply hand the whole order to the broker and say: go for it! Give me the best market price you can over the day. But what does this actually mean? And how did we know if a good job was done?

Easy peasy, readers may think. Track Nintendo’s share price from the open until the close and see if the broker did better than the average? But it’s not so simple. Hidden from view is the number of shares traded at each price.

For example, I may look at the intraday share price chart and be miffed that the broker missed a few hours of very low prices before lunch and lumped me with the much higher prices later on. My entry point may be way above the average.

But this could mask the fact that few shares were traded early in the day. A thousand at 8.30am. Two thousand at 8.32am. Only in the afternoon did sellers of Nintendo place big orders in the tens of thousands against which mine were matched.

Hence the metric institutional investors use when scrutinising trades is a volume-weighted average price. Did we beat VWAP overnight on those Sony trades, I would ask Gara of a morning? Nah, Daiwa screwed us, he might reply.

Now, we ran pension plans. The good or bad execution of trades was not as important in the long run as picking the right stocks at the right time. But with average annual percentage returns of mid-single digits, basis points here and there add up.

And when markets are extremely volatile — as they were in April — we’re talking more than basis points. Take my purchase last month of a certain equity exchange traded fund. It was during the Orange Crash and I was hoping to pick the bottom.

I placed the order on the evening of April 10. Trouble was, I turned out to be too clairvoyant for my own good. Hours later Donald Trump did his first major U-turn on tariffs and stock prices all over the world cheered.

My ETF immediately jumped 6.6 per cent the following morning to £25.19 per share, before declining 3.7 per cent throughout the day to close at £24.28. I received the price of £25.15 — almost the high of the day. Where was Gara when I needed him?

Luckily, I’ve made a nice return regardless. But had my order been executed at £24.50, I would currently be up 6.7 per cent — almost 3 percentage points more than the return I have, thanks to a crappy entry price.

That difference is pretty much half the average annual real return of the S&P 500 over the past century, one of the best-performing stock markets ever. You don’t want to be giving that kind of money away regularly.

What can readers do about it, though? In the example above, the online broker I use to run my self-managed pension cannot be blamed. Overnight orders are always executed at the open on a strict queue.

I could have limited my order of course, perhaps requesting that shares were only bought below a certain level. And none of this would have been a problem if I was trading during market open hours, where orders are routed to wherever prices are most attractive.

Other options are available too. So-called algorithmic orders adjust their pricing in relation to what is going on in the market. Investors can also set time limits on their trades and other fancy tricks depending on the asset class.

But all of these approaches to trading assume that I know which direction prices are going to move on a given day. And that is impossible ex ante — for the same reason the majority of active managers underperform the index.

When it comes to the exact prices you trade at, in other words, you win some and lose some. Best not to worry too much and go with a broker whose execution costs — that is the fee they charge per trade — is the cheapest. 

The author is a former portfolio manager. Email: stuart.kirk@ft.com; X: @stuartkirk__