Time for an asset allocation revolution

By Rayann Huang | February 26, 2010 | Last updated on February 26, 2010
9 min read

Risk is a four-letter word nobody likes, and portfolio managers are no exception.

However, the emergence of several new asset allocation methods that move away from the fixed mix strategy gives cause to see risk in a new light. These risk-focused methods can lead to better portfolio choices and thereby provide better protection during market crashes.

Markets are in constant flux, risk-focused theorists assert, so asset mixes should shift periodically to maintain the portfolio’s exposure to risk, i.e. volatility level. A risk-based approach to asset allocation is an active style of management that breaks away from the traditional buy, hold and rebalance strategy. It adapts asset mixes to evolving market conditions and rebalances them back to the portfolio’s target risk level.

Risk-based approach

Although the risk-based approach is not entirely new to the institutional world, it is relatively unknown among retail investment firms. In Canada it has been adopted by large institutional investors such as the Ontario Teachers’ Pension Plan, the Canadian Pension Plan Investment Board and Caisse de depot et placement de Quebec.

As with most quantitative methodologies, the challenge in executing the risk-based approach effectively is gathering the right data and analyzing it correctly. Across the board everyone will do it differently.

Toronto-based, Pur Investing Inc. is one of the few Canadian investment firms offering risk-based portfolio management to retail investors. Their risk-budgeting approach manages exposure to market volatility by constantly rebalancing portfolio assets to a target risk level. To do this, the firm’s managers continually assess historical market volatility, rather than historical returns. Ioulia Tretiakova, director of quantitative strategies, says the decision to focus on historical volatility is based on common sense, because past returns are no help in forecasting future returns, whereas past volatility can reliably be used to predict future volatility.

“The autocorrelation [the correlation of a variable’s future values with its past values over a successive period] of monthly volatility of the S&P 500 [standard deviation calculated using daily returns during the month] is 67% from January 1950 to December 2009,” says Tretiakova. “In contrast, the autocorrelation for returns for the same period is 5%.”

Pur’s president, Mark Yamada, explains the constant rebalancing to risk levels ensures the client is not taking on more risk than desired.

“If you look at risk of the U.S. market as represented by the S&P 500 over time, the risk of that market [as measured by the 21-day standard deviation] is 15% and yet last year it spiked at 83%,” he says. “So to maintain the same asset mix for a client when the risk of the market has moved from 15% to 83% is possibly irresponsible. What professional managers should have been doing as the risk flew very rapidly to 83% is taking risk out of the client’s portfolio to maintain a 15% risk in that portfolio. With risk in the stock market at 83%, you probably own none of it [equities].”

Despite the ongoing shifts to clients’ asset allocation, amendments to their investment policy statements are not required because their portfolios would continually be rebalanced to their fixed-risk levels.

Fixing the mix

Stephen Fricker, CFP, principal of Westdale Financial Services in Hamilton, has developed his Portfolio Recovery Strategy Program (PRSP) to give clients an alternative to the fixed-mix strategy. After suffering through the recent market crisis, many of his clients who are nearing retirement said they weren’t prepared for another roller coaster ride again and wanted a different option.

The PRSP is not a risk-based approach per se. It uses quantitative risk measurements, economic fundamentals and historical returns to align asset mixes to changing market conditions. In a market with high volatility, clients’ portfolios will have a larger proportion of bond funds to equities. Conversely, when volatility is moderate, bond funds will gradually be traded for equity funds.

His program offers four different investment models as guidelines for account design and ongoing management. Each holds a different proportion of fixed income to equities as its name suggests: Income, Conservative, Moderate and Aggressive. He also has sub- categories for each model that he’s named based on different market stages: Recession 1, Recession 2, Recovery 1, Recover 2, and several others.

Investors are matched with a model based on their investment profiles and as the market evolves their asset allocations will transition into different pre-formulated sub-categories of asset mixes with different ratios of equities funds to bond funds and cash.

For example, if the client’s account is based on the Moderate investment model, in a Recession 2 sub-category his or her portfolio would be 100% in fixed income and then when the markets started to recover, the asset mix would move into a Recovery 1 sub- category that consists of 48% fixed income, 25% growth funds and 28% special growth funds.

In a few circumstances, the client may be moved into an entirely different investment portfolio model. For example, last year, Fricker advised many of his clients in the Conservative model to adopt the Income model.

Fricker primarily looks at Price/Equity ratios and economic fundamentals as a proxy for market volatility, and as signals to transition the client into a different asset mix.

“As the market prices go up slowly, then that P/E goes up and up. So if you want to keep their level of aggression the same they need to scale it back,” says Fricker. “The same thing with conservative investors; if their highest point of equity investment is at 40%, when the market is at its lowest, then keeping at 40% when the P/E is high, is disrespecting a conservative risk tolerance.”

Setting goals

The transition to a different asset mix can also be precipitated by a goal. If clients have achieved a specific investment objective they may no longer need to take on the same levels of risk.

Whether the clients are transitioning to a new asset mix or investment model, the risk level will change, as noted by the three-year standard deviation score value Fricker discloses on the written proposal to make the change. However, he says, by changing the asset mix to adapt to current market conditions, overall exposure to downside risk sometimes remains more or less the same.

“There is a big change if a person is switching from a conservative portfolio with a standard deviation score of 4.4% to an income portfolio with a score of 11.5%, but this is in recognition that the market has dropped 25% to 30% in value,” says Fricker. “So our ability to go in at 11.5% at a price that is 30% lower comes up at the same level of risk for the client who goes in at 4.3% with prices 30% higher.”

If the price is low and the client’s exposure is high, this starts to balance with an asset mix that has lower equities and enters the market when P/E ratios are high. In that way Fricker can stay in a fairly narrow corridor of volatility risk in an account. “By doing this, I help maintain the client’s risk tolerance,” he says.

The move to a different asset mix or portfolio model is preceded by a proposal that shows shifts in asset class weighting and the portfolio funds’ distribution of risk [low to high]. Clients are required to sign the proposal if they agree with it, and this document becomes their updated IPS.

Updates to the KYC are required if there is more than a 10% change to the weighting of an asset class and risk distribution. For example, if the proposal raises equities holdings to 50% from 35% and the weighting of medium risk funds jump to 30% from 15%, updates to the KYC would be required.

In terms of returns, Tretiakova says her firm’s approach builds a floor on maximum losses that a portfolio can have (what the investor can comfortably lose) and maintains the model that returns are proportionate to risk. Noteworthy from their return data is that with better downside protection higher returns and lower risk are not necessarily mutually exclusive.

The test of a hypothetical backdated investment portfolio that was constantly rebalanced for a 10-year period from January 20, 2000 to January 20, 2010 showed better returns than the S&P500. Constantly rebalancing the portfolio to a conservative risk level of 11.4% produced annualized return of 0.3%. Not bad compared to the S&P500 that had a moderately aggressive risk level of 21.3% and had an annualized return of -2.3%.

During a shorter period, October 31, 2007 to February 29, 2009, a medium aggressive portfolio that was constantly rebalanced to its level, had a lower loss of -21%, compared to the S&P’s 50.9% decline.

Yamada notes advisors often say the average return for U.S. stocks in the long term has been 8%. So, over 15 years clients think they can expect a return of 8% compounded for 15 years. “Well that’s really hard to deliver,” he says. “But if you know what your risk is on the downside you’re always going to get the risk of the market, so if there is an upside you’ll get it. Any portfolio managers out there will say, ‘Gee if you can save money in down markets, you’re much better off than shooting the lights out in the good market.’ ”

Risk Management

Fricker manages the risk return relationship differently. He looks at the clients’ overall investment objectives and advises them on the level of risk each should take given the economic climate and how the portfolio has performed.

For example, clients who had double-digit returns from 2002 to 2006 were advised to lower their risk levels in 2008 because the portfolio gains from the bull market helped to make up for some of the losses and lower returns from the bear market.

“We made 55% to 60% over that four-to-five year period. So the point is that we made over budget; why subject yourself to risk?” says Fricker.

In January 2008, he moved a client into a portfolio comprised of 100% fixed income; as a result her return for the year was -8.2%. But he started to get her back into the market early in 2009 and client incurred some losses. Yet, Fricker explains that by taking a conservative posture, she made 14.8% so far this year and recovered. “An 8.2% loss means she needs 10% to recover and she’s at 14% . . . she’s way ahead of everyone else,” he says.

One of the reasons most people stay loyal to the traditional fixed-mix strategy is they believe it has merit in the long term [DASH] a theory that stems from the Modern Portfolio Theory and the Capital Asset Pricing Model.

Yamada does not dispute this theory, but he says the challenge it faces is defining what is meant by long-term. “If you ask portfolio managers their ideas of long-term, you will get anywhere from 18 months to ten years.”

Perhaps some of the redeeming qualities of the asset-based process are its simplicity and lower management costs. Managing a portfolio’s exposure to risk requires ongoing monitoring and a lot more math than most people are comfortable with, says Tretiakova. Furthermore, with frequent shifts in asset mixes there is the risk of trading at inopportune periods, which is why both Yamada and Fricker contend their methods do not time the market and instead use a cautious, disciplined process.

“The risk is usually in buying back when the market has already started the recovery and getting out of the market too early – for example, when volatility started rising before the tech wreck broke out.,” says Yamada. “So the approach tends to err on the side of caution.

Fricker adds trading risks do not diminish the value of his program. “Sure there is risk in the model. I fully understand the difficulty in implementing it, but contrast that to the risk of doing nothing.”

Then there are the costs: the higher trading expenses and larger tax consequences that come from more buying and selling.

Fricker provides full disclosure of the switch cost and tax consequences. The switch costs range from $125 to $275 based on the client’s overall household market value. By providing explicit reporting, he and his clients can weigh costs and benefits of moving assets. Moreover, it helps for tax planning by encouraging clients to look for tax-loss selling opportunities in the future.

“It is not easy to reconcile mix shifts with the tax cost, but I would submit that anybody who saved 25%-to-40% of their equity portfolio last year would gladly have paid the tax,” says Yamada. “The rebalancing overlay considers the tax consequences of rebalancing decisions and rebalancing trades occur only when the net economic benefit of the trade, including risk reduction, transaction costs and taxes, is positive. “So the turnover will be a function of many variables, including transaction cost structure, portfolio size, and the tax status of the portfolio.”

Rayann Huang is a Toronto-based financial writer.

Rayann Huang