Investment Commentary > Past performance is not a reliable indicator

Past performance is not a reliable indicator

Past performance is not a reliable indicator of future performance.

I love this disclaimer. It’s more than a disclaimer – it’s become an investment cliché. A phrase I think most people are at least vaguely familiar with and have some (accurate) understanding of what it means.

And yet, what’s remarkable is that investors constantly think and act as if recent performance is their best indicator of future performance.

The psychologists like to use the term “recency bias” to describe this sort of behaviour – the reality that many individual’s perception of future events is excessively influenced by recent events.

Check out this wonderful example. An investor survey released recently raised a lot of eyebrows – certainly mine. It was carried out by Natixis Investment Managers. Their “2021 Natixis Global Survey of Individual Investors” surveyed 8,550 investors in 24 countries. All “wealthy” investors with a meaningful amount of net assets. With respect to anticipated future investment returns, US investors reported that they expect returns of 17.5% annually in coming years.

17.5%. Per year. 17.5%…

I mean… wow…

I couldn’t resist playing around with this and what it means. So if you take a lousy $500,000 and invest it for 5 years, you will end up with $1.12 million. And then I guess you can just kick back and live off the return, which by then would be close to $200,000 p.a.

Investors do have some good support for their optimism. Earnings estimates for the S&P500 are also projecting a stellar few years to come. Sadly, there’s some things where past performance is indeed a reliable indicator of future performance and we all know the earnings forecasting game, right? Let’s recap…

At any point there’s a bunch of “forward earnings estimates” for individual companies as well as the S&P500 as a whole. These estimates are done by the analyst community – principally the major banks and brokerages firms – the “sell side”. Various data companies aggregate the individual projections and at any point there’s “consensus forward estimates”, being an average of the projections.

For the firms producing the estimates, analyst research reports and earnings estimates are basically marketing documents.

As I’ve discussed in the past, these guys can’t be “bearish”. When your business is selling investment products and services, being negative is entirely non-commercial. Think about that in relation to any other business and you will see what I mean.

I just looked out the window and saw a bus go by with a big Harvey Norman advertisement on the side – let’s use them as an example;

Imagine walking into Harvey Norman with thoughts on updating your TV or fridge or coffee machine or smartphone. They happen to have a sale on – what luck! Now, imagine you start talking to the sales rep. She says to you…

“Well, okay… let me tell you… there’s some really exciting developments taking place in those products… we think there’s going to be some really great new models released soon and rumour has it they might even be priced cheaper than the current models…”

Now of course she doesn’t say that! The perfect product for you is one of the ones they have in stock right now. The best time to buy is right now. Investment firms sell investment products, services like capital raisings and investment management services. They will not turn customers away. That’s the case whether they are a major investment bank and the customer is a major corporate looking to do a multibillion-dollar deal or if they are a small retail stockbroker talking to a prospective client about investing $10,000 for the first time. If the advisor wants to succeed in his job, that customer will be told that they came to the right place at the right time.

That’s not to say that the major investment firms know a change in market direction is in the winds… We know financial markets are incredibly difficult to predict (and they can hide behind that fact). But I guarantee you plenty of sell-side analysts prepare research they don’t really believe in.

Anyhow I’m digressing a little here… back to the earnings estimates game…

Analyst consensus estimates are often followed for about the next 2 years. Right now, we’ve got estimates as to how this calendar year will close out and also how 2022 will turn out. Given results are published quarterly, there’s generally quarterly estimates and then yearly estimates, being the sum of the quarterly estimates.

Dr Ed Yardeni does us all a big favour in monitoring and publishing data on this sort of stuff. A visit to his website is highly recommended (but be warned – you might get lost for hours in a sea of fascinating charts).

He likes to call the little worms created by the changes in consensus estimates over time “earnings squiggles”. Here’s the set fort the last decade or so:

Just to fully explain what’s going on here, let’s focus on the one labelled “13” (representing the 2013 year).

Notice it (and all of them) are about 2 years in length.
The ’13 begins in 2011 – being when consensus estimates for the 2013 year were first tracked.
It starts at around 125 on the Y-axis – this is the initial consensus estimate for 2013 earnings.
It then drops down… and down… and down some more…
Finishing at around 110.

Analysts are forever tweaking their estimates. That’s okay right? As “the event” gets closer there’s greater clarity as to the final outcome and surely they are allowed to change their estimates. I mean, it’s a bit like forecasting the weather a week from Saturday – as the date approaches, it gets much easier to accurately predict the result.

Sure – its fair that analysts continue to refine their predictions as new information becomes available. But here’s the thing… look at the squiggles. They practically all slope down to the right… in other words, the end actual earnings are almost always below (often well below) initial estimates.

This chart for a longer duration illustrates the phenomenon even better. Indeed, history shows that initial analyst estimates overshoot the actual number at least 80% of the time

It’s the same in individual stocks – that’s the game! Analysts throw a number out there for 1 to 2 years in the future and along with it comes a valuation and price target for the stock.

Then there’s this almost choregraphed routine whereby company managers massage down analyst expectations over the quarters, resulting in them reducing their earnings estimate.

Then… pow!

We’re in the midst of another US reporting season and, as usual, most companies are beating consensus estimates. Of course they are. It is literally rigged so that companies almost always beat estimates. That’s not conspiracy theory nonsense – it’s a well-known game that dates back to at least the ‘90’s.

How can we use this information?

Well firstly, we use this knowledge by being incredibly sceptical about consensus earnings estimates – understanding that they almost always prove excessive optimistic. But there are some other valuable takeaways from this. Notice that analysts have not forecast any of the recent recessions/downturns – massive “misses” versus their initial estimates.

But notice after a recession, earnings usually bounce back quite strongly and within a few years they have tended to surpass the prior peak. Look at the 2000 to 2003 and 2007 to 2010 periods. What’s also true is that, at least over this last 20 year period, the market has tended to have a good period following a recession. The earnings rebound and “upward revisions” is a logical driver in this.

But here’s the rather unique thing about this most recent Covid recession/recovery period. The recovery began with valuations already high. Some investors are looking for a glorious multi-year uptrend based on historical precent. I commend them for looking at history, but unfortunately it is actually different this time.

To digress a moment… just look at that chart. Wow… Investors think they can generate 17.5% returns in a market that’s already done that? I realise that’s simplistic, but mean reversion is very real in financial markets and this market has some mean “mean-reverting” to do at some point…

The Drivers:

Revenues must reliably follow the economy. That makes sense, right? In a developed nation, circa 70% of the economy is consumer spending. Consumer spending is corporate revenues – both goods (first chart below) and services (second chart). And when we have a look through history that is exactly what we see:

Notice how nominal GDP has averaged in the order of 5% for a long time. Again, we can observe that corporate revenues have averaged a very similar number.

But corporate earnings have grown in excess of revenues in recent times. How’s that possible?

Well, the main reason is that corporate profit margins have grown to be well above long term trend.

There are some logical mathematical reasons for this – a large government deficit helps! But history shows that profit margins are as mean-reverting as the market.

So we’re at a point where profit margins are beyond historical norm and analyst consensus estimates of future earnings are for significant continued earnings growth, implying further margin expansion. Sounds a bit optimistic to say the least.

Choose your own reality.

What I thought I would do is prepare for you two alternative market analysis pieces – two “alternative realities”, if you will.

The first is the type of thing you are hearing from sell-side analysts – the talking heads on TV, much of the internet and your friendly stockbroker/financial adviser. The second is one that you will get from many historically-informed buy-side investment managers – people such as… well, me…

Option 1:

The rebound in the economy post-Covid continues to impress. As we’re seeing, corporate profits are also bouncing back strongly with the majority of companies beating analyst estimates. Its only logical that shares have performed strongly.

Looking forward, the economic expansion should continue quite strongly and the outlook for corporate profits is therefore solid. When we look at forward estimates, 2022 earnings are around $215 for the S&P500 and there’s still some scope for further upside revisions here. The market certainly has had a strong 12 months. Despite this, with a forward PE of around 20 times, valuations remain quite reasonable. When we compare this to interest rates and bond yields, stocks continue to look attractive and with central banks remaining very accommodative, it is reasonable to expect share markets will continue to perform well in this environment.

Option 2:

If there is one irrefutable truth to investing, it is the reality that the higher the price today, the lower the prospective returns. Share markets have certainly performed incredibly strongly for the past 12+ years. Whilst many investors want to believe that performance is “normal”, understand that all those past gains lower future returns.

When it comes to valuing assets, value is derived from the very long stream of cash flows that stand to be delivered to the holder of that asset. Despite what many analysts contend, value is not simply based on next year’s earnings or the year after that. Indeed, there is little correlation to any single year earnings and share price.

It follows that any reliable valuation model should seek to value the asset against long-term cash flows. When you have a play around with different valuation models, we also observe that models based around revenues have better predictive powers than models based around profits. This is logical given profit margins have proven to “normalise” over time, just like valuations.

In terms of valuation models, I’m frequently citing Dr John Hussman’s work and several of his excellent models he generously shares with the world. For a change of pace, today let’s look at some other historically-reliable models, courtesy of the team at Advisor Perspectives.

The chart below shows four models they track:

Consistent with other reliable models, these too are indicating that the market is about as over-valued as it has ever been.

It is critical to remember that high valuations are not a trigger for a market decline. If the market fell every time valuations became elevated, the market would never be able to work its way to extreme valuations like it has today. Valuations reliably predict longer-term returns not short-term returns.

The uncomfortable reality is that extremely valued markets tend to resolve via significant unwinds – look at 2008 and 2000 for guidance. Investors seem to be taking solace in the idea of “… yeah but they were big events… why would this market implode like that?”

The apparent absence of a catalyst should not provide comfort. History shows that many an overvalued market unwind happens without a catalyst. The “reason” for the falls is very often only identified and attributed after the fact.

Despite what many analysts claim, shares are extremely valued and wholly unattractive at this point. Whilst poor returns on a 5 to 10 year horizon are practically guaranteed, that’s not to say that further gains don’t lie around the corner. But as historically-informed investors we need to accept that valuations will revert towards mean at some point and nobody will ring a bell at the top.

I’ll leave you to choose your own reality…

This document contains information which is the copyright of Aviator Capital Pty Ltd (AFSL 432803) or relevant third party. Any views expressed in this transmission are those of the individual, except where the individual specifically states them to be the views of Aviator Capital Pty Ltd. Except as required by law, Aviator Capital Pty Ltd does not represent, warrant and/or guarantee that the integrity of this document has been maintained nor is free of errors, interception or interference. You should not copy, disclose or distribute this document without the authority of Aviator Capital Pty Ltd. Aviator Capital Pty Ltd does not accept any liability for any investment decisions made on the basis of this information. This information is intended to provide general information only, without taking into account any particular person’s objectives, financial situation, taxation or needs. It does not constitute financial advice and should not be taken as such. Aviator Capital Pty Ltd urges you to obtain professional advice before proceeding with any financial investment.

Register your interest in this Fund

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Ut elit tellus, luctus nec ullamcorper mattis, pulvinar dapibus leo.