Portfolio construction is as much an art as it is a science. It’s a science because, thanks to modern portfolio theory, advisors can construct efficient frontiers securing the best risk-reward tradeoff. It’s an art because no two clients will have the same risk profile – or better, risk appetite.

ETFs can simplify the process or they can make it more complicated. They can simplify, because they are more transparent than mutual funds: the holdings are available every day. Risk statistics are also more reliable but they can complicate the process because, as more and more ETFs are launched, filling almost every nook and cranny in the marketplace, finding the right fit involves some work.

Building an ETF portfolio requires some basic steps. Step One may be to take a look at the existing portfolio of actively managed funds. If they follow a benchmark, there might be a suitable – and cheaper – ETF equivalent. Step One could also be to develop an ETF protfolio from scratch. In either case, it’s an asset allocation decision: which classes of assets, and in what combination, are clients comfortable with: how much in bonds, in Canadian and in non-Canadian equities.

Not all ETFs are alike – nor are they always suitable for a portfolio, says Pat Chiefalo, director of Derivatives & Structured Products Research at National Bank Financial Markets, as he spoke at the Exchange Trade Forum sponsored by Radius Financial Education.

Key considerations in ranking the suitability of specific ETFs for a portfolio are liquidity – not only trading volume but bid-ask spreads. Larger ETFs tend to be more liquid. Then, there’s performance or better: tracking error. Some ETFs, whether because of currency hedging contracts or only holding a sample of the stocks, have returns that deviate significantly from the underlying index.

Next comes costs: not all ETFs have bargain-basement management fees. They may cover the same asset class but do so in different ways. Finally, there is diversification: do the top 10 holdings overwhelm the rest of the ETF. That makes for a concentrated bet.

Still, NBF has created model portfolios – along the familiar classification of income, conservative, balanced, growth and maximum growth investors. That’s the asset allocation piece. Finding the best ETFs for a proposed asset allocation is another matter.

Before he begins ranking anything Chiefal asks…Are these products we want to consider for the portfolio? If there’s any doubt they get kicked. Once they go into the ranking, they are all generally pretty good.

But what should that asset allocation do? For Cheifalo, it may require a bit of tweaking to match investor expectations. But Ioulia Tretiakova, director of quantitative strategies at PUR Investing takes another step, building on the ease with which ETFs can be used to maintain a constant risk budget.

She uses the example of past S&P 500 peaks. A buy-and-hold investor would have seen the risk characteristics change over time – the return is overwhelmed by the risk, as happened during the tech bubble in the late 1990s.

With a fixed risk budget for each investor – essentially, how much of a loss they are willing to withstand each year – Tretiakova can switch between bonds and equities. She insists this is not market timing. Rather, it’s to maintain a constant risk profile.

“We are decreasing our equity exposure as volatility is increasing in order to maintain constant risk and we would be going more conservative if volatility is increasing,” she says. “How do we know what volatility will be like in the future? The amazing thing about volatility is that it is highly predictable. It is highly predictable because it is persistent, which means that past levels of volatility are indicative of of what is going to happen in the future.”

Chiefalo takes a different approach. First comes the asset allocation then comes “tweaking” if it’s not living up to expectations. “You set up your initial asset allocation and you see how it performs over time,” he says, giving an example. “I really had a lot of volatility with my equity underlying as opposed to my fixed income. So, maybe I want to rethink my initial asset allocation and tweak it to take into account what I’m seeing over time. It’s more kind of an iterative process rather than something that’s dynamically changing over time. Mine really doesn’t take the view of more active bets: a lot more Canada or a lot more bonds.”

By contrast, Pur Investing tackles risk directly, by building portfolios with each investors “risk number” in mind. That risk number is essentially the maximum annual drawdown an investor is willing to accept. A risk number of 15 means a 15% loss in one year.

But its more complex than that, because, through an extensive questionnaire, Pur tries two combine two levels of analysis. The objective part is what level of risk is necessary to obtain the “aspirational return”. The subjective part is what level of risk the client is comfortable with. These can conflict.

“We will try to communicate to the client that by restricting their desired level of risk they are limiting their return potential, but that obviously depends on the client and our communications skills,” says Tretiakova.

Sleep well at night or reap the most: ETFs can help, but there’s no magic bullet.