Our approach to constructing investment portfolios is best described as value-aware: investing in high-quality assets when they are attractively priced.
Quality determines whether an asset or asset class is worth owning. Price determines when to invest and how much.
Value-aware investing requires assessing these factors separately including accurately estimating an assets intrinsic value to determine whether the current market price is attractive.
Why valuation matters: the two engines of investment returns
Asset prices do not move in a straight line with their fundamentals. They often can overshoot in both directions before reverting towards their long-term trend. This is known as mean reversion. It is one of the more reliable patterns in markets, although its timing is impossible to predict precisely.
This matters because valuation influences future returns. If you buy a high-quality asset below its long-term valuation trend, your return will likely be driven by two sources. The first is growth in its underlying economic value, driven by factors such as earnings, rents or population growth – this all relates to the asset’s inherent qualities or fundamentals. The second is the uplift from the market repricing the asset’s valuation back towards its long-term average.
Conversely, a high-quality asset bought at an expensive price may still produce a positive return. However, growth in its underlying value may be partly or entirely offset, as its valuation reverts lower – towards its long-term trend.
That is why both quality and price matter.
Why property valuation matters most when you buy
With property, the most important valuation decision is the one you make when you buy.
Once you own an investment-grade asset, your job is to patiently and diligently hold it through multiple market cycles, ideally for several decades.
The relationship between price and intrinsic value matters more when investing in property than shares for two reasons.
First, property is a lumpy, illiquid asset with high transaction costs. A share portfolio can be built gradually by investing in many tranches over time. With property, you must invest a substantial amount of capital at a single entry point. Therefore, the price you pay matters enormously.
Secondly, while price tends to drive returns in the short term, quality ultimately does the heavy lifting over the long term. You must hold property long enough to allow that quality to translate into growth. As a rule of thumb, the final third of a multi-decade holding period will generate more than half of the asset’s total capital growth. Therefore, selling during a flat or below-trend period because the asset appears temporarily overvalued risks forfeiting the most valuable phase of the investment.
How to assess whether a property market is undervalued
The best way to assess whether property in a particular market is attractively priced is to compare current values with long-term trends, as I did recently in this blog. This involves comparing median property prices with where they would be if they had grown in line with their long-term average.
Of course, this analysis only provides a city-wide average. Valuations can vary significantly between locations and property types within the same city.
Investors can therefore consider several other measures to build a broader body of evidence that supports or challenges the view that a market is undervalued:
- Rental yields relative to their long-term average. Above-average yields may indicate that rents have grown faster than property values, suggesting prices have been subdued for some time.
- Relative values between property types. For example, comparing the historical relationship between house and apartment prices may indicate that apartments are undervalued relative to houses, or vice versa.
- Values relative to other capital cities. These relationships should also be assessed over long periods. For example, median house prices in Brisbane and Adelaide have recently exceeded Melbourne for the first time, which is historically unusual.
- Replacement cost. This involves estimating the value of the underlying land plus the current value of the improvements. If an existing property can be purchased for less than its replacement cost, that may be another sign that it is undervalued.
The challenge is that property valuation and evaluation are highly subjective. Unlike the share market, there is no substantial body of daily pricing data, and no two properties are identical.
Therefore, no single measure will provide a definitive answer. The objective is to combine several data points to build a sufficiently strong investment case.
How to assess whether a share market index is attractively priced
Shares require a more ongoing assessment of value because investors typically invest new capital progressively over time. Fortunately, there is a broader range of valuation metrics available, including price-to-earnings, free cash flow, price-to-book and dividend yield. Each provides a different perspective but also has different limitations.
Price-to-earnings ratios: are current earnings sustainable?
The price-to-earnings ratio compares a company’s share price with its reported earnings per share. It is the most widely used valuation metric, largely because it has the longest and most standardised history, making comparisons with long-term averages relatively straightforward.
Its main weakness is the “E”. Although company-specific accounting distortions tend to be diluted across a diversified index, they do not disappear. Reported earnings remain an accounting measure rather than cash flow, and aggregate earnings can be distorted by write-downs, impairments and other one-off items.
Therefore, the important question is not simply, “What is the index’s P/E ratio?” but, “Are its earnings maintainable?” At an index level, this means assessing whether aggregate profit margins and earnings are sustainable, particularly given the economic cycle, changing business conditions and the index’s sector composition.
The second complication is whether to use trailing or forward earnings. A trailing P/E is based on earnings that have already been reported, making it more reliable. However, it is backward-looking and may be misleading if profit margins or economic conditions are changing. A forward P/E may better reflect what investors are currently paying for, but it relies on analyst forecasts. While aggregation reduces company-specific forecasting errors, analysts can still be collectively too optimistic, particularly late in an economic cycle.
I typically use trailing P/E as the anchor for comparison with long-term averages because it provides the longest and most consistent data series. I then use forward P/E as an indication of how the market is pricing near-term changes. A large gap between the two can sometimes be a red flag.
As a broad rule of thumb, a trailing P/E ratio of around 18 times is often considered normal for a developed-market index. However, the appropriate multiple will also depend on factors such as interest rates, expected growth and the composition of the index.
Overall, P/E is a useful first-pass valuation measure because historical comparison data is abundant, but it should never be relied upon in isolation.
Free cash flow: the most honest measure of value?
As the saying goes, “Revenue is vanity, profit is sanity and cash flow is reality.” In simple terms, the intrinsic value of a share market index is the present value of the future cash flows generated by its constituent businesses. If there’s little future expected free cash flow, there’s little value.
Free cash flow, typically calculated as operating cash flow less capital expenditure, is arguably a more economically honest measure than reported earnings. Cash received and spent is generally harder to manipulate than accrual-based profit. At an index level, free cash flow yield measures the aggregate cash generated by the underlying businesses relative to the price investors are paying for them.
However, interpreting free cash flow across an index is complicated because different sectors have vastly different capital requirements. Headline free cash flow also deducts both maintenance expenditure, which is necessary to sustain existing earnings, and growth expenditure, which is intended to increase future earnings. Although both reduce cash flow today, they have very different economic purposes.
This distinction is particularly relevant to the recent AI boom. Many of the world’s largest technology companies are spending extraordinary amounts on data centres, semiconductors, energy infrastructure and the broader rollout of AI. This expenditure reduces free cash flow today but, if it generates substantial future earnings, it could create significant long-term value.
Of course, we should not automatically assume that all AI spending is genuine growth investment. Technology changes quickly, computing infrastructure can become obsolete, and competition may force companies to keep investing simply to maintain their market position. What is described as growth expenditure today may become tomorrow’s maintenance expenditure. If the expected AI revenue and productivity gains do not materialise, some of this investment may ultimately be destroyed rather than create shareholder value.
This also makes historical index comparisons less reliable. Large technology companies were once relatively capital-light businesses, but AI is making parts of the sector far more capital intensive. Given their substantial index weightings, this shift can materially affect aggregate free cash flow. Free cash flow is also less meaningful for some sectors, particularly financial companies.
Unlike P/E ratios, there is no widely accepted long-term average free-cash-flow multiple for developed-market indexes. As a broad guide, these markets have typically been valued at equivalent to around 17 to 25 times free cash flow. However, differences in definitions, index composition and aggregation methods mean this should be treated as an indicative range rather than a reliable historical benchmark.
In practice, I regard free cash flow as the most economically honest of all the measures, but also the most difficult to benchmark confidently across markets and over long periods. Therefore, it works best as a sanity check on other index valuation measures rather than as a stand-alone valuation tool.
Price-to-book: what accounting value leaves out
Price-to-book compares a company’s share price with the accounting value of its net assets.
Dimensional, one of the pioneers of factor-based investing, has traditionally used P/B as its primary value measure. The rationale is not that P/B predicts returns better than other valuation metrics. Dimensional tested it against 9 alternatives, including price-to-earnings and price-to-cashflow, and found no reliable difference in expected returns. Instead, its advantage is that book value tends to be much more stable than earnings or cash flow. This reduces portfolio turnover and provides a cleaner value signal, without blending in as much exposure to profitability.
The main limitation is what book value excludes. Accounting standards generally do not recognise internally generated assets such as brands, research and development, intellectual property and human capital on the balance sheet. Therefore, P/B tends to be more useful for asset-heavy businesses such as banks, industrial companies and resource producers, where book value provides a reasonable indication of the assets supporting the share price.
It is less reliable for asset-light businesses such as software companies, service businesses and consumer brands. These companies may possess substantial economic value that does not appear on their balance sheets, causing P/B to make them look more expensive than they really are. This also reduces its usefulness when assessing indices heavily weighted towards intangible-rich businesses.
Overall, P/B is a useful portfolio-construction and screening tool. It is most reliable when applied to capital-intensive, asset-heavy markets or sectors, but should not be used in isolation when assessing markets dominated by intangible-rich businesses. A “normal” P/B ratio is around 3x.
Dividend yield: a useful valuation cross-check
Dividend yield is best used as a cross-check rather than a primary valuation signal. If P/E and P/B ratios suggest an index is attractively priced, dividend yield should broadly support that conclusion by sitting above its own long-term average. In simple terms, a lower market price should produce a higher dividend yield in percentage terms.
Similarly, when an index’s trailing dividend yield falls below its historical range, it provides another piece of evidence, alongside P/E and P/B, that the market may be pricing in more optimism than usual. Conversely, an above-average yield may support the case that an index is undervalued, provided it reflects sustainable dividends rather than temporarily inflated payouts or deteriorating earnings.
Over recent decades, broad market-cap-weighted international developed-market indices have generally traded at dividend yields of around 2% to 2.5% p.a., compared with approximately 4% to 4.5% p.a. for the S&P/ASX 200, excluding franking credits.
Which index valuation measures matter most?
None of these four metrics should be used in isolation, but they do not deserve equal consideration. My ranking, from most to least reliance:
- Free cash flow is the most economically honest signal, because cash is harder to manipulate than accounting earnings, but it is the hardest to benchmark confidently over long periods due to limited standardised historical data. Best used as a quality check on the others.
- Price-to-earnings has the richest long-run historical data to benchmark against, which makes it the most practical starting point, provided the earnings number itself is interrogated for quality and maintainability first.
- Price-to-book is a reliable corroborating tool, particularly for asset-heavy markets, but should be weighted down for indexes dominated by intangible-rich businesses.
- Dividend yield is purely a corroborating check. It should move in the same direction as the other three; when it does not, or when it is being propped up by an elevated payout ratio, that divergence is itself useful information.
The strongest case for an index being attractively priced is when several of these measures agree, each read against its own long-run average, rather than any single ratio taken as a standalone verdict.
Valuing an index in practice: the FTSE 100
Let’s use the UK’s FTSE 100 Index as an example.
| Valuation Metric | Current | Long-term average |
| Trailing P/E Ratio | 11.6x | 15-16x |
| Forward P/E Ratio | 11.6x | 13-14x |
| Price-to-Book (P/B) | 2.3x | 1.7-1.8x |
| Dividend Yield | 3.0% | 3.4-3.8% |
The FTSE 100 index appears inexpensive relative to current earnings, but expensive relative to its book value and, to a lesser extent, its dividend yield.
These measures are not necessarily contradictory. The combination of a low P/E and high P/B implies that the index is generating a much higher return on book value than it has historically. In simple terms, its constituent companies are producing more profit from each dollar of book value. This may reflect elevated profitability across the banking, energy and resources sectors, together with a lower accounting equity base.
However, its below-average dividend yield suggests that a smaller proportion of those earnings is being distributed as dividends. This may reflect greater earnings retention.
This is a good example of why no single valuation metric should be considered in isolation. Looking at several measures provides a more complete picture and highlights the assumptions that require closer examination.
Where to find reliable index valuation data
It is important to treat online valuation data with caution because it can quickly become outdated, and different providers may calculate the same metric differently.
However, there are several reasonably reliable sources. Australian ETF providers tend to publish monthly fact sheets that include portfolio characteristics and valuation metrics. Index providers such as S&P Dow Jones, FTSE Russell, etc., also commonly publish equivalent data on their websites.
Finally, AI can be useful for conducting deeper research across multiple sources. However, any findings should be corroborated, fact-checked and checked for both the calculation methodology and effective date.
Why index quality matters as much as valuation
When constructing an ETF portfolio, the quality of the underlying indices and their methodologies is critical. For more on maximising portfolio quality, refer to my previous blogs on alternative indexing methodologies and ETF portfolio construction.
The takeaway: quality first, then price
Attractive pricing is not about buying something simply because it looks cheap. It is about buying a high-quality asset below the value implied by its long-term fundamentals.
Property rewards making a sound entry decision and then holding patiently for decades. Shares reward combining evidence-based index selection with valuation awareness, rather than diversifying blindly.
