Why forecasting returns is not black and white

Tim Farrelly
Share on facebook
Share on twitter
Share on linkedin
Share on email

‘I don’t see how financial planners, or anyone, can set investment strategies without forecasting returns’ – Tim Farrelly

Institutional investors tend to have constraints on when they can, versus should, make significant changes to asset allocation – much more so than financial planners, according to Tim Farrelly, a veteran advisor to the advisors. But this works in favour of his business, he believes. He spoke with David Chaplin about the processes behind his forecasts.

Tim Farrelly has a colourful view of the world. Right now, for example, Farrelly sees US equities as yellow; local shares, a kind of blue; and Australian government bonds daubed in red.

But his colour-coded asset classes are more science than conceptual art. The four-colour palette is simply the visual output of an underlying system developed by Farrelly to forecast asset class returns over 10-year periods.

In the model, marketed under the farrelly’s Investment Strategy brand: green equals cheap; blue represents fair value; yellow is fully valued, and; red represents over-priced.

There’s more to Farrelly’s process, of course, than mix and matching colours. He says the 10-year return forecasts are based on current asset metrics – such as P/E ratios for shares – and a raft of macro-economic data.

Farrelly applies the forecasting system to a range of asset classes including shares, bonds, hedge funds (“I think of them as an asset class”) and more granular sectors.

“Anything that I think is reasonably forecastable, I’ll do it,” he says. “Some things are easy to forecast – like Australian banks – and others, such as the resources sector, are really hard, so we won’t do them.”

But while the analysis may be robust enough there’s nothing particularly complex or revolutionary about Farrelly’s forecasts – except for the fact that forecasting is something of a dirty word in an industry more comfortable with the past.

“I don’t see how financial planners, or anyone, can set investment strategies without forecasting returns,” Farrelly says. “Just relying on historical returns is incredibly misleading.”

In particular, he says ‘efficient market’ disciples use historical data to support “blatant nonsense” while engaging in “convoluted arguments” to explain why reality diverges from theory.

“At the other end, you’ve got investors who don’t know what they believe – they’re worse than the efficient market guys,” Farrelly says. “We often see it at conferences where the audience vote on whether they agree with speakers. Even where there are two speakers who had strongly opposing views, about 80 per cent of the audience sometimes agrees with both.”

However, he says there is a growing “middle group” of investors who have strong beliefs but will consistently test their assumptions and change them when presented with the evidence.

According to Farrelly, institutional investors don’t typically belong to this latter group. He says while institutions may know when asset prices are getting out of whack, “business and career risks” limit the action they can take.

“It’s good for my business,” Farrelly says. “Financial planners don’t have those constraints and they can take a position that is in the best interests of their clients.

“Most institutions would rather sell out of an asset class too late than too early. But if financial advisers go early, they go early and clients don’t begrudge that if the message is communicated well.”

And Farrelly emphasises he produces forecasts, not predictions, which his clients use to set strategies as opposed to time markets.

For instance, if US equity P/Es push the sector into Farrelly’s red zone “then I’d tell my clients to sell US shares”.

“But the model doesn’t say when a correction will happen – only whether something is fair value or expensive,” he says. “So I’d say [if the sector went red] don’t sell out of US equities fast – do it gradually over eighteen months to two years. Ideally, the prices would go up for two years then collapse but even if the prices fell before then, this approach is a lot better than doing nothing.

“Advisers know they’re doing the right thing in reducing exposure to overpriced assets but it’s nothing to do with timing, because we just don’t know when a turnaround will come.  Just that we know that it is very likely to happen, and the longer it takes before it does the worse the fall will be.”

In fact, Farrelly says his forecasting model works better during bubbles than business-as-usual markets.

“But that’s the most important time when you can show dramatic over-valuations,” he says. “The rest of the time the results are not so clear but it doesn’t matter so much.”

Farrelly’s model has real-life bubble data to back up this claim. His quarterly forecasts date back to 2004 when he launched the business, with the model throwing out respectable returns since, especially pre and through the GFC.

“I didn’t see the GFC coming but prices were pretty frothy leading up to it,” he says. “I’d say most of my clients managed to sell down about 25 per cent of their risky holdings before the GFC hit.

“Since the GFC, [the model] returns have pretty much matched the market.”

Farrelly deals with a wide range of financial advisory groups in Australia and New Zealand, who may implement his asset forecasts in different ways.

“I’m agnostic [on implementation],” he says. “I believe that active managers can add value but they’re very difficult to find. I can’t do it.”

More importantly, Farrelly encourages his clients to write down their investment philosophies and regularly question those beliefs.

“Are those beliefs still valid? What has happened that may invalidate them,” he says, citing the example of his preference for Australian fixed income exposure via term deposits rather than government bond funds.

“Until the GFC it was true that Australian government bond funds offered better returns then TDs,” Farrelly says. “But since then TDs have returned 1-2 per cent better for what is essentially a government-guaranteed investment. That’s a clear case for why you’d change your philosophy, others are more subtle.”

His own beliefs have also changed substantially since his first contact with the investment industry in 1976. At the time, Farrelly was hired for background clerical duties at the investment tipsheet run by industry pioneer, Jim Cowan.

His association with Cowan continued while studying for an engineering degree he completed in 1980. Farrelly never got to build bridges, though, fetching up at Macquarie Bank in 1988 for what would be an almost 15-year career with institution.

Over that time, he filled numerous roles in product design, asset allocation and finally, heading up Macquarie’s retail distribution arm.

It was during his final years at Macquarie that Farrelly was inspired by the writings of Vanguard legend, John Bogle, on asset class valuations.

While the Bogle ideas proved difficult to instill into the Macquarie culture, Farrelly later pursued the thoughts at the University of Technology Sydney, writing a paper on different approaches to asset allocation.

“I was surprised that no-one was doing anything different,” he says. “Everyone was doing strategic asset allocation, which was obviously flawed.”

More than a decade down the track, Farrelly estimates he has about $10 billion in ‘funds under influence’, nailing his colours to the mast.

Share on facebook
Share on twitter
Share on linkedin
Share on email