The DIY Financial Advisor 3 – Decision Framework and Asset Allocation

Today’s post is our third visit to the book – The DIY Financial Advisor (from the team at Alpha Architect).

DIY Financial Advisor

DIY Financial Advisor is a book from the team at Alpha Architect, an asset manager and a consultancy to family offices.

The thesis of the book is that the DIY Investor can make long-term decisions designed to maximise risk-adjusted performance after fees and taxes.

• In contrast, the fund manager has short-term incentives which conflict with this goal.

The themes of the book are that you can do better than the “experts”, but that you need to avoid psychological traps and stick to an investing framework.

In the first part of the book, AA made a convincing case that a systematic DIY investor – armed with a model – can beat the experts:

• Experts are self interested, susceptible to the same behavioural biases as the rest of us, and rely on stories rather than facts.

Humans are prone to poor decision-making across a broad range of situations.

Or as Dan Ariely puts it:

The human mind is predictably irrational.

---

Systematic decision-making (which uses models) outperforms discretionary decision making (using experts).

The (erroneous) idea that experts can outperform relies on three incorrect assumptions:

1. That qualitative information increases forecast accuracy.
2. That more information increases forecast accuracy.
3. That experience and intuition increase forecast accuracy.

Humans have three key problems:

1. Inconsistency
2. Overconfidence
3. Reliance on stories

Going DIY

Experts don’t want you to follow an evidence-based systematic process.

• It prevents them from charging you a fat fee.

So they use four tactics to persuade you to stick with them:

1. Fear – if you go it alone, you might lose all your money.
   • “We will keep you safe.”
2. Greed – you won’t have access to our secret ways.
   • “These opportunities are reserved for our best clients.”
3. Complexity – a simple model can’t beat the market.
   • “We have armies of MBAs and PhDs, researching how to beat the market.”
4. Relationships – stick with the firm that’s served you for decades.
   • “You trust us, don’t you? Let me buy you lunch.”

---

AA provide a list of ten reasons why not to use an expert (eg. a financial adviser):

1. They charge too much
2. You can’t tell what they are charging
3. They try to sell you “company-approved” products
4. They are aggressive in their recommendations
5. They are constantly changing your investments
6. Their investment strategy is too complicated (for you to understand)
   • This may be coupled with early (and good) advice on simpler topics, so that you first evaluate them as competent
7. They don’t listen to your opinion
   • And if they do agree, perhaps they are simply catering to your wishes to gain trust initially
8. They take more notice of their clients with larger portfolios
9. They won’t take responsibility for their mistakes
10. Their investments have poor (risk-adjusted) returns

People seem to be hard-wired to trust smiley young people in suits with lots of qualifications and good advice on simple topics.

• Do your best to resist.

Decision framework

AA’s decision framework is called FACTS:

1. Fees – should be as low as possible for the asset class / strategy in question.
   • Asset manager and investor incentives should be aligned where possible.
   • Closet indexers who overcharge for a simple strategy should be avoided.
2. Access – liquidity is desirable.
   • Avoid lock-ups, gates and redemption fees.
   • You might need to make an exception in return for tax breaks (eg. VCTs / EIS).
   • Prefer closed-end funds for illiquid assets (eg. property).
3. Complexity – you need to decide on a level that’s appropriate for you
   • For example, I’m happy to use a version of volatility optimization (hand-crafting) to determine allocations between asset classes.
   • But I would not use this approach to produce a 20-stock portfolio from the FTSE-100 (because the historical volatilities and covariances there would not be stable enough into the future).
   • Always look at what you are getting (higher returns, lower volatility) in exchange for additional complexity.
   • Beware of data-fitting to small sample sizes.
   • How stable is the system, and how does it handle risk-management?
4. Taxes – tax efficiency is crucial but often over-looked
   • You can’t spend pre-tax money.
5. Search – the costs of identifying how to implement a strategy (finding the funds / managers) need to be subtracted from the returns of that strategy.
   • And don’t forget the costs of monitoring the funds / managers that you choose, and of dealing with fund and manager turnover.

---

In addition to this decision framework, you need to understand some portfolio management fundamentals:

1. Asset allocation
2. Risk management
3. Security selection

Asset allocation

AA’s discussion of asset allocation begins with Markowitz and modern portfolio theory.

This has many shortcomings – notably that volatilities and covariances are not stable, and that returns do not follow a normal distribution – but the underlying message remains valid:

• Diversification is a free lunch.

Combining assets with similar but uncorrelated returns produces a less volatile portfolio.

• AA warn against mean-variance optimisation using small sample sizes.

They also note that equal weighting of factors / assets can often perform as well as more sophisticated models.

• We agree, and the hand-crafting approach we use includes a lot of equal weighting.

---

AA use an asset model with three classes:

1. Equities
2. Bonds
3. Real Assets (real estate and commodities)

Our own model replaces Real Assets with Alternatives, and sub-divides this class into three:

• Equity-like Alternatives
• Bond-like Alternatives
• True Alternatives

Risk management

AA note that a moving average strategy (trend-following) can protect investors from large drawdowns.

• Using a 200-day MA leads to better returns than buy-and hold,  both in absolute and risk-adjusted terms.

It’s refreshing to see this version of market timing explained as a risk management measure, rather than as a means of out-performance.

Security selection

AA use this heading to discuss market anomalies or factors:

• value
• momentum

As usual, human behavioural biases are at the root of them.

A simple asset allocation model

In chapter 6, AA provide a simple asset allocation model – something that I tend to file under the heading of a Lazy Portfolio.

• I think that Lazy Portfolios – those made up from a few funds, often four or fewer – are attractive to investors with small portfolios.

But as my portfolio grew, I was happy (indeed eager) to take on more diversification (or as AA would characterise it, more complexity).

Fans of Lazy Portfolios also need to decide how to handle home bias (to the UK) and away bias (many Lazy Portfolios are US-biased).

---

My base passive portfolio currently includes 36 asset classes.

• I fully accept that this is too much for most people to cope with, especially at first.

But I firmly believe that market cap indexing is sub-optimal

• And the simplest and safest way to improve it is to sub-divide major components into sub-funds, and allocate equally between them.

For investors with less than say £100K in their portfolio, 16 to 20 assets would be sufficient.

• There’s a big gap below those numbers, straight down to Lazy Portfolios with fewer than ten assets.

---

AA begin with a quite from Einstein stressing the courage required to make things simpler.

• I agree, but I would note that his famously said that things should be made as simple as possible – but no simpler.

They move on to a 2009 paper from DeMiguel et al which compares a basic equal-weight allocation to many other complex strategies.

• Equal weight does well, and the out of sample results for the more finely-tuned allocations are not good.

But these finely-tuned models are not used by any private investor I know.

Changing numbers

To critique the MPT model, AA look at the changing returns, volatilities and covariances of three assets:

1. S&P 500
2. Hedge funds
3. 10-year Treasuries

The numbers are different when comparing two periods:

• 1927 – 1960, and
• 1961 – 2013.

More importantly, using these sets of numbers produces very different portfolios, and (in the pre-1960 case) some extreme weights.

---

This is to be expected, but for me it doesn’t invalidate the diversification benefits identified within MPT, or the use of lots of asset classes:

• Unless you have some way of predicting the numbers for the future (the out of sample period), you might as well assume that those for the past (the in sample period) will persist.

Hand-crafting is in any case a compromise process which protects us from extreme allocations.

• It uses robust long-term correlations, and the simplifying assumption that Sharpe ratios between assets cannot be reliably determined.

---

For me, the number of assets in a portfolio should be determined by portfolio size (or strictly, by a minimum deal size).

• If you have enough money, why not use more assets?

More assets will also protect against the risk of extreme allocations.

Equal weights

AA compare an equal weight allocation to the min var portfolio (what I would call the max SR portfolio) and the tangency portfolio.

• Equal weights produces good returns (better than the other two) but this is not a real-world comparison.

The max SR portfolio is low-return, since it has a low allocation to equities (which are very volatile).

• Thus in the real world, the big asset allocation is how much equities?

For return maximising investors, the equity allocation will be around 75%.

• For conservative investors, 60% or even 40% equities would be appropriate.

These are the portfolios than need to be compared with equal weights.

• And the analysis needs to involve more than three assets.

Endowments

To make the discussion more realistic, AA move on to a consideration of the Yale University endowment fund.

• Large university funds have become famous over the past couple of decades for producing good returns from multi-asset portfolios.

The Yale allocation in 2015 was:

• 31% private equity
• 20% absolute return (hedge funds)
• 19% equity
  • 13% international, 6% US
• 17% real estate
• 8% natural resources (real assets / commodities)
• 5% bonds and cash

AA say that this would be hard to implement for private investors, but I would not allocate this way even if it were easy.

• Less than 20% in stocks, and more than 75% in alternatives is pretty weird to me.

I like all those asset classes, but not in those proportions.

---

AA also have qualms, but they reduce the portfolio back to three assets (five or six in my book):

1. Stocks
   • Foreign, domestic and private
2. Bonds
   • US
3. Real assets
   • Real estate and commodities

Lazy portfolios

In the next section, AA look at some lazy portfolios.

First, Buffett:

• 90% S&P 500
• 10% cash

Next, Bogle:

• 50% S&P 500
• 50% bond index (we’ll be generous and assume a global allocation)

Third, the trad 60/40:

• 60% domestic equities (US)
• 40% bonds (international government bonds?)

Fourth, Bernstein:

• 25% domestic equity (US)
• 25% international equity
• 25% small cap (again, lets assume global)
• 25% bonds

Fifth, Swenson (from Yale):

• 30% domestic equity (US)
• 15% international developed equity
• 5% emerging markets equity
• 20% real estate
• 15% inflation-protected bonds
• 15% US Treasuries

---

AA compared the last three over the period from 1979 to 2014, and declare the result to be a three-way tie.

• 60/40 did best, but this was a particularly good period for those two assets.

AA recommend more diversification as a general approach.

• But they also conclude that “advanced” asset allocation may be unimportant.

They may well be right, but their real point is that it is not worth paying for.

• Today, DIY investors can access dozens of asset classes at bargain-basement prices, so this is not as much a concern as it has been in previous periods.

You can safely take on as many asset classes as you which.

• But you’ll need to decide for yourself how many that is – DYOR.

Ivy 5

The portfolio that AA decide on is the Ivy 5 from Meb Faber, which uses equal weights:

• 20% US stocks (S&P 500)
• 20% International stocks
• 20% real estate
• 20% commodities
• 20% bonds (7 – 10 year government bonds)

This is not bad if you only have five funds.

• But personally, I wouldn’t be comfortable with a 40% target allocation to stocks.

---

That’s it for today.

• The two chapters we’ve looked at have been much more meaty than the four we examined in the previous articles in this series.

AA make a convincing case for intelligent lazy portfolios, but have not quite-convinced me to abandon my multi-asset portfolio.

• For me, hand crafting is a good compromise of robustness and simplicity.

That I end up buying 36 ETFs rather than five is an irrelevant detail to me.

• And with less cash, I might use as few as 16.

In the next article, we’ll look at risk management and security selection in more detail.

• Until next time.