What should we be doing in asset allocation?

In my last article I discussed what I believe are the limitations to the current approach to asset allocation of much of the financial planning (and funds management) industry. In this article I attempt to provide the basis for an alternative.

My main criticism was that the asset allocation process generally adopted had essentially become the provision five to six ‘pick a box’ standard passive allocations using conventional assets. This approach was too simplistic, overly constrained and often ineffective in meeting client objectives and managing risk.

How important is asset allocation anyway? Some people believe asset allocation constitutes 91.5 per cent of returns. Are such precise figures misleading? It depends. If you only use passive exposure to index funds, then asset allocation constitutes 100 per cent of returns. If you use much more active, skill-based strategies, then asset allocation’s contribution may be as low as 20 to 30 per cent. Indeed, if you construct a portfolio more from the bottom-up focusing on actual investment selection and more skill-based strategies, asset allocation becomes less important and is primarily an outcome of the process rather than the key top-down driver.

This is the first step towards an alternative asset allocation approach. In downplaying asset allocation by adopting a more bottom-up approach, one requires greater flexibility at the top-down asset allocation level. As few as four broad classes — equities, property, cash/fixed interest and alternatives — may be appropriate. Decisions on components of these — local versus overseas, fixed interest versus cash, listed versus unlisted and so on — are decisions that should be made at an underlying investment strategy level and depending on the types of managers and vehicles that are available.

I cannot see the logic of the conventional approach, which commits to exposure across all regions, sectors and styles without even considering the fundamentals behind these or whether good managers or vehicles can be found.

The second step is to cut back the number of model asset allocations. Having as many as five to six allocations is little incentive to manage each asset allocation with a proactive focus.

After all, once the strategic asset allocations are set (usually based on 15 to 20 years of history and what competitors are doing), adjusting a client’s asset allocation is usually achieved by simply moving them into a higher or lower equity mix. If the number of allocations is cut to two or three there is clear differentiation and a greater focus on structuring each mix to meet specific objectives.

The third step is to widen the asset allocation ranges. This is to allow for the flexibility to manage exposure at the investment strategy level and to allow for significant moves. Such moves do not have to be frequent, but the value added in moves of one to two per cent hardly justifies the trouble. If you are two per cent overweight something that outperforms by five per cent over the 12 months you have added all of 0.1 per cent to return. Is this worth it?

Of course, the general response to moving away from passive asset allocation is that no-one can time markets. This line is one of the biggest cop-outs going around investment markets. It gives planners an excuse to do nothing and to use fund managers that do next to nothing. In fact, anyone involved in the investment business is automatically involved in a range of timing decisions. Choosing which of the standard five to six asset allocations to use, timing the actual investing of a client’s money and choosing the style of fund managers used are all, in part, timing decisions, particularly in the mind of the client.

In any case, I am not arguing for frequent short-term asset allocation changes. Short-term timing of markets is difficult and something only a few talented and disciplined managers and hedge funds can do.

In fact, momentum rather than valuations or fundamentals is the key driver over such short time frames, which is probably why economists tend to make the worst short-term market timers. However, this is one reason why including an active hedge fund/managed futures component can make sense in a portfolio.

However, to complement this, a medium-long term approach to asset allocation is feasible for those with discipline, a valuation focus and common sense. The aim is to be willing to adjust exposure to markets at extremes, recognising that this could result in reducing exposure to those markets performing well and increasing exposure to those markets doing badly. In the short-term, this can result in underperformance versus competitors.

However, if asset allocation is seen primarily as a risk management strategy, then such periods of underperformance can be tolerated and long-term outperformance should be achieved.

Why do most fund managers tend to add so little value in asset allocation? I believe this is because they firstly attempt short-term timing without the appropriate skills, process or discipline, and secondly because they attempt asset allocation almost on a casual basis without committing adequate resources.

In practice, a sound asset allocation approach may mean only one or two big decisions every few years. But how can your average fund manager justify employing resources to make such infrequent decisions? As a result, asset allocation gets delegated to an investment committee whose individuals are mainly focused on other things, who meet and think about asset allocation at set times only and who are under pressure to add value on a monthly or quarterly basis. It is not surprising that the results are disappointing.

Asset allocation should be more forward looking with a lot less emphasis on how a particular asset mix has performed in the past and more emphasis on return prospects and downside risk in the future.

Valuations should be the key driver, although long-term business cycles and sentiment should also be considered.

The key questions include where are price to earnings ratios (PEs) and other valuations compared to history, what happens to returns if they revert to historical averages, when was the last economic/earnings downturn/upturn, is it a bull or bear market environment and how exposed to the markets are investors.

While this may not give definitive answers to the appropriate asset allocation, it is better than relying on recent history as most models do.

Asset allocation today should also encompass some of the alternative asset areas. The case for alternative assets such as hedge funds, commodities and private equity is very strong from an investment and risk perspective, yet few asset allocations encompass these assets to any significant degree.

Again, if we are not talking at least 10 to 20 per cent of a portfolio, it is no surprise that planners ask, “why bother”?

Despite the questions being raised about the conventional approach and the poor performance of many portfolios, it is likely that most planners and investors will be slow to move away from it.

One practical approach is to have a core portfolio component that takes a conventional passive asset allocation approach, but to supplement this with a component that has a more forward-looking and innovative asset allocation, including alternative assets.

There are no easy answers when it comes to asset allocation, but it is time to at least partly move away from the backward looking quasi-science that drives today’s approach to focus on what asset allocation should really be about — diversification across a range of sensible investment ideas whose forward looking fundamentals (and not just their history) stack up. Isn’t that what investors are paying for?