Anyone considering hiring a financial adviser asks themselves two questions: Will I be better off with the advice than I would have been without it? And does it matter which adviser I choose?
The industry answers the first question with confidence. Vanguard publishes an estimate suggesting that advice can increase annual net returns by "up to, or even exceeding, 3%", a large improvement in performance.1 Russell Investments puts the value of an Australian adviser at 5.5% per annum.2 These figures are frequently cited and shape both clients' expectations and the way advisers present themselves.
The few published research findings are less clear-cut. Investigating whether advisers make a difference presents a number of difficulties when designing a research project. That may sound technical, but it simply means that it is difficult to reliably distinguish between the contribution of financial advice to personal wealth and other reasons why people become wealthier, such as differing attitudes, work ethic, relationship status, the number and age of children and other such factors. The technical term from epidemiology refers to whether the research succeeds in sufficiently ruling out the influence of confounding factors.
People who seek advice differ from those who do not, even before the advice is given. Most studies are unable to distinguish between the effects of the advice itself and the characteristics of those who chose to seek it. Studies that are able to track individual advisers find that they differ greatly from one another, and not always to the benefit of their clients.
The second question has been raised in the field of medicine. There, studies measure how much of the variation in treatment outcomes can be attributed to the individual medical doctor, after taking the patient's own risk and all other known information about the doctor and the patient into account. The proportion is sometimes negligible and sometimes substantial.3,4 To date, no study has addressed this question in relation to financial advisers.
This page sets out what has been measured, by whom and with what limitations. It then discusses what the findings mean for someone choosing an adviser, as well as for an adviser evaluating their own work. It is aimed at interested clients or potential clients, advisers and researchers.
On 17 August 2026, the practice conducted a search of the published literature. The search initially covered Australia, then the United States, the United Kingdom and Europe. It included journal articles, working papers, dissertations, reports from regulatory authorities and materials published by companies.
Companies' own estimates are presented alongside published critiques of them. The author's own experiences are identified as such.
The Office for National Statistics' Wealth and Assets Survey covers Great Britain. In the 2014 to 2016 survey wave, people who reported receiving advice were, on average, older and £274,000 wealthier than those who did not.5 They were 15% more likely to hold a university degree and 15% more likely to be married. Furthermore, they were more likely to receive lump-sum payments both before and after receiving advice.5 A comparison of the two groups therefore measures both the difference between the groups and the possible impact of the advice.
Statistical methods can attempt to eliminate the influence of each of these factors in various ways, but there may be other, unknown factors that also differ between the two groups. The influence of these factors, as they are unknown, cannot be mathematically eliminated. The most common method of minimising or eliminating the influence of even unknown factors is to randomly assign individuals to either the group receiving advice or the group not receiving advice. This is known as a randomised controlled trial. While this may work in clinical trials of a medicine, it is not a realistic scenario in this case.
A published systematic review searched journals and grey literature up to January 2021 for all studies the reviewers could find on the value of personal financial advice.6 The final sample comprised 286 records: 170 journal articles, with the remainder consisting of grey literature such as doctoral theses or non-peer-reviewed publications.
Broken down by region, 134 originated from North America, 99 from Europe and 35 from Australia and New Zealand.6 Of the literature reviewed, 80% define value as the financial advantages of advice. Only 16% of the studies examine long-term value based on clients' financial records.6 The review reports that "few studies include matched-pair samples or account for potential selection bias from those who seek advice, which calls into question the robustness of these results".6 A matched-pair sample compares individuals who received advice with those who did not, selected to be similar in as many characteristics as possible, such as age, gender, relationship status, income and assets. This is a research design (also known as a case-control design) that is useful for observational studies and can yield valuable insights; however, as previously mentioned, there may also be further systematic differences between the two groups that are unknown.
The Financial Conduct Authority tracked the same British survey respondents across five rounds of data collection, from 2010 to 2020.5 It compared the change in wealth of 540 people who received advice with that of 1,778 people who did not, while taking into account the characteristics recorded in the survey. Two years after receiving advice, the wealth of those who had received advice was 10.2% higher.5 This is the only estimate whose 95% confidence interval excludes the value zero. This is referred to as a statistically significant result and is often an indication of an actual effect.
Subsequent estimates remain positive but widen, and the estimate falls to 1.6% after eight years.5 After eight years, the interval ranges from −16.5% to +19.7%. A confidence interval that includes zero generally means that no conclusions can be drawn from these data. This is particularly true when the confidence interval is as wide and as symmetrical as it is here. The uncertainty stems from the extent to which changes in wealth vary. Between the survey waves, the median change in wealth for individuals without advice was 7.6%, while the mean change stood at 728%.5 A small number of very large changes, such as lump sums or growth from a very small base, pulls the mean up.
In the note, the estimate was also calculated excluding individuals who had received a lump sum, such as an inheritance, a gift or a one-off pension payment. The correlation was then "not clearly different from zero".5 The authors advise caution when interpreting the magnitudes or extrapolating them to other contexts. The survey captures advice received over the preceding two years, meaning that a first appointment and a long-standing relationship give the same answer.
The US panel "The Asset and Health Dynamics among the Oldest Old" asked in 1993: "Do you have a financial advisor who helps make decisions?"7 Among respondents aged 60 and over, 13.09% answered "yes" to this question. A PhD student at Texas Tech University used this single response to develop models of net worth across seven subsequent survey waves from 1995 to 2008. The models were adjusted for net worth at the start of the survey, income, educational attainment, cognitive abilities and demographic factors.7 Having a financial adviser in 1993 was positively associated with subsequent net worth, particularly more than a decade later.
Three characteristics of the data influence this result. Firstly, average net worth in 1993 was $143,497 for those without a financial adviser and $357,686 for those with one, approximately 2.5 times as much.7 Secondly, changes in inflation-adjusted net worth differed significantly between the groups up to 1995, 1998 and 2000, but not over longer periods. Thirdly, the number of respondents fell from 6,183 in 1995 to 1,846 in 2008.7 In a re-analysis focusing exclusively on those still alive in 2008, having had an adviser in 1993 was significant for net worth only in 2004, 2006 and 2008.7 The question regarding an adviser was asked in this longitudinal study only in 1993, but not in subsequent years.
In Vanguard's study, the total value of advice is put at "up to, or even exceeding, 3% in net returns".1 It states that such opportunities arise "intermittently". The February 2019 issue stated: "We do not believe this potential 3% improvement can be expected annually; rather, it is likely to be very irregular."8 On Vanguard Australia's consumer website in December 2021, the figure is quoted without these caveats: "financial advice improved net returns by 3 per cent."9
A critique by Derek Tharp on Kitces.com raises three points.10 One component, savings on fund costs, assumes that the adviser reallocates the client's investments into the cheapest few per cent of funds. The figure is calculated before the adviser's fee is deducted. And there is no counterfactual, "since we don't necessarily know how the client would have acted in the absence of an advisor".10 In the author's view, this criticism is itself questionable, as index funds were not yet a widely used investment option at that time, meaning that people without an adviser might be less likely to choose index funds than clients with an adviser. Kitces.com itself sells memberships, professional development courses, conferences and training programmes to advisers.10 Therefore, index funds may not be Kitces's preferred option.
Russell publishes a figure for Australia every year. The 2026 edition puts it at 5.5%: 1.5% for asset allocation, 2.8% for behavioural coaching and 1.2% for tax matters.2 The 2025 edition stated 5.6%, and the 2023 edition 5.9%, of which 3.4% was attributable to behavioural coaching.11,12
The behavioural component represents a twenty-year gap between investors and the market. The 2023 report takes into account the monthly inflows and outflows from US funds, as reported by the Investment Company Institute.12 It calculates the single rate of return that matches these inflows and outflows: 6.65% per annum. This is compared with the 10.04% per annum that an S&P 500 portfolio would have achieved had it remained unchanged from 1 July 2003 to 30 June 2023.12 The difference of 3.39 percentage points corresponds to the behavioural component. The report is stamped "FINANCIAL PROFESSIONAL USE ONLY" and states that it "must not be shared with the end investor".2
A Canadian study reported that households which had had an adviser for fifteen years or more held 2.73 times as much financial wealth as comparable households without an adviser.13 A later survey put this figure at 3.9 times. The study's own document reveals that both surveys were commissioned by a financial firm.13 No direct comparisons are made between households with and without advisers. First, the same survey is used to estimate how likely it is that each household has an adviser. Subsequently, households predicted to have one are compared with households predicted not to.13
The idea is simple. A fund's published return relates to a dollar that remains invested for the entire period. The average dollar actually invested achieved a different, usually lower, return, as investors buy after good years and sell after bad years. The difference is the investor-return gap. DALBAR's annual study puts this at around four percentage points per year for investors in US equity funds and sells the study to advisers so that they can pass it on to their clients.14 Morningstar's "Mind the Gap" puts it at 1.2 points per year for US funds and 0.40 points for Australian funds.15 The behavioural coaching component in the figures from Vanguard and Russell is this gap, which is regarded as what an adviser prevents.12 Morningstar itself interprets the small Australian gap as a sign of "more holistic financial advice" in Australia, which is an interpretation of its own figures, not a measurement.15 Neither of the two studies records whether an investor had an adviser. The gap also fluctuates depending on the timing of contributions and fund costs, even if no one makes a poor decision. A peer-reviewed re-examination of Morningstar's sample put the part due to poor timing at 0.10 points a year.16 The Active and passive page presents both the studies and the published critiques, in the sections on DALBAR and Mind the Gap.
This is a summary of the claims made by financial advisers on their websites, along with the justifications they provide for them. In August and September 2026, the accessible websites and Financial Services Guides of Australian advisers were archived and analysed. LinkedIn pages are not included, as LinkedIn's terms of use prohibit the downloading of large numbers of pages. A language model, "Jev" by TypeSafe, version 1.13.0, read 8,967 pages from 2,220 adviser websites, covering 4,277 advisers, and assessed each page against five kinds of clearly stated claim.17
| What the page claims | Share of websites |
|---|---|
| The adviser is good: expert, experienced, trusted, without making any comparisons | 84% |
| Clients receive a benefit other than returns: peace of mind, security, freedom | 77% |
| The adviser is the best or has received an award or ranking | 23% |
| The adviser is better than other advisers, firms or banks | 14% |
| The adviser is independent | 9% |
| The adviser achieves better investment returns | 5% |
Almost every website claims that the adviser is good, and most promise a benefit other than returns. "Peace of mind" appears on 543 of the 2,220 websites.17 Claims about better returns are rare. Of the 110 websites making such a claim, 39 quote a figure and 18 cite a source for it.17 Some of these figures are repeated estimates from the industry itself: one website quotes 5.2% per annum from an industry study, another 5.6%.17
Websites that compare themselves with other advisers rarely name a competitor. The comparison is made with "most advisers", large companies, institutions or banks. Websites that claim to be the best usually refer to a list or an award: "top 100 financial advisers" according to a published list, "adviser of the year".17
The reasons advisers give for choosing them usually relate to the individual. Long-standing experience leads, appearing on 75% of websites. This is followed by personalised service at 70%, qualifications at 67% and values or ethics at 53%. External recommendations are less common: client testimonials at 31%, awards at 20%. A fee model, such as "fee-only" or a flat fee, is the rarest reason, at 8%.17
"Independent" is a term whose use is restricted under the Corporations Act, and an adviser who is not independent must state this in the Financial Services Guide. This statement was found on 27% of the websites.17 One in nine websites (11%) uses wording that claims independence: "independent financial advice", "an independent adviser", "we are independent". Three in ten of these websites also include a statement indicating that they are not independent. Around 7% of websites claim independence without denying it anywhere on the site. A further 13% state that they are independently owned, privately owned or unaffiliated, which constitutes a different claim. Advisers who hold their own licence are most likely to claim independence (19%), while those with the largest licences are least likely to do so (6%).17
There is, however, one caveat: the study only covers those 28% of current advisers whose own websites were found and verified.17
The bodies that award certifications make similar claims about their holders. Most describe their certification as the "gold standard" or a mark of excellence. None claim that holders achieve better returns.17 One Australian organisation, the FAAA, offers evidence for a client benefit of CFP certification: an annual online survey it commissions, the Value of Advice research.18,19
This survey assesses how around 1,200 respondents rate their own lives and finances. Clients of CFP professionals rate their situation most highly. The reports provide no information on how the respondents were recruited, how a respondent was identified as a client of a CFP professional, or how many such respondents there were.18,19 They do not take into account who selects an adviser, nor do they measure investment outcomes. In the only edition in which group sizes were specified (2022), 52% of advised respondents reported a CFP adviser.20 Around one in four advisers holds this designation.17 In the author's experience, few clients know whether their adviser holds the CFP designation. The full analysis of the survey can be found on the page about the websites study.17
Four Canadian mutual fund distributors provided the researchers with their client and transaction data.21 The main analyses are based on the three datasets with a complete historical record: 5,920 advisers and 581,044 clients, from January 1999 to June 2012. These advisers are permitted to sell only mutual funds, may only carry out transactions with the client's consent, and are remunerated through commissions from the funds.21
Thirty-three measured characteristics of each client explain 13% of the variation in clients' share of growth assets.21 These measured characteristics include age, income, wealth and responses to the risk questionnaire. When the adviser is taken into account, a further 19%, totalling 32%, of the variation in growth assets can be explained.
Switching from a conservative adviser at the 25th percentile to a more aggressive one at the 75th percentile is associated with a 20 percentage point increase in growth assets.21 This corresponds to half the range between clients with the lowest and highest risk tolerance, who held 40% and 80% in growth assets respectively. Consequently, an adviser's personal aggressiveness in selecting the proportion of growth assets (shares and property) for all their clients is a strong predictor, even stronger than a client's demographic characteristics, of how aggressively their clients invest.
An adviser's own portfolio is one of the strongest predictors of a client's portfolio.21 The average client paid more than 2.7% per annum in fees. A later study puts the advisers' own net returns at around −3% per annum.22 As so-called passive or index funds, mainly due to their lower costs, have historically achieved higher returns than active funds overall, the advisers' own portfolios lagged just as far behind the passive benchmarks as those of their clients; yet despite this objective evidence, they continued to hold on to such portfolios even after leaving the industry.22
Two of the authors of the working paper received financial support from Canadian financial firms.21 A "switcher" analysis, which tracks clients who change advisers, supports the interpretation that advisers build similar portfolios for their clients. An alternative and quite possible further hypothesis is that clients who choose the same adviser are similar to one another.21
At one large Swiss retail bank, every trade was classified as "advisory" or "independent".23 In the year following the purchase, the advisory purchases generated a return that was 3.0 percentage points lower than the return on the same client's own purchases. This figure compares each purchase with the return that a share with the corresponding risk would have been expected to achieve on the Swiss market. The study presents the difference in three ways: in terms of raw returns without taking risk into account, the recommended purchases lagged by 1.1 points, a difference that is statistically indistinguishable from zero. Taking into account the risk of each share compared with the Swiss market, the shortfall was 3.0 points. When global market risk, company size, value and recent price performance were also taken into account, the shortfall was 1.7 points.23 The latter two figures are statistically significant. The entire shortfall arose from purchases made following contact from the adviser: these yielded returns that were 3.3 points lower. Purchases initiated by the client themselves and subsequently reviewed showed no difference.23 The advisers were able to reduce some of the clients' behavioural biases, but not to a sufficient extent to offset the difference in performance.23
In a university pension scheme in the United States, brokers were initially available to new participants but were no longer available later on.24 Clients of brokers achieved lower returns than the matched portfolios from the scheme's target-date funds (the low-cost, passive option often intended for disengaged clients), after taking market risk into account. They also achieved lower returns per unit of volatility, the so-called risk-adjusted returns, despite bearing a similar level of risk, and paid average brokerage fees of 0.90% per annum.24 The report also highlights the other side of the story. Before the scheme offered target-date funds, the default investment for members who did not make a choice was a money market fund holding only short-term deposits. Participants who went through a broker were much less likely to remain in this default investment.24 As soon as target-date funds became the default investment, far more members remained in this default investment: 44% compared with 16%.24
In 2011, ASIC conducted a shadow-shopping study of retirement advice.25 Real consumers sought advice, and ASIC analysts assessed the advice using a written evaluation form. Of 64 advice cases, 58% were rated "adequate", 39% "poor" and 3% "good".25 Of the same 64 participants, 86% rated the quality of their advice as good, and 81% had a high level of trust in it.25 Only three participants rated their advice as "poor". ASIC concluded that "many people have difficulty in objectively assessing the quality of advice they receive".25 All 25 examples rated as "poor", representing 39% of the sample, failed to meet the statutory standard for advice, at that time section 945A of the Corporations Act. ASIC referred each individual case to its Misconduct and Breach Reporting team for further investigation.25
In a US audit study, trained researchers posed as clients and presented the portfolios assigned to them.26 The advisers encouraged returns chasing and recommended actively managed funds with higher fees. They did so even when the researcher presented a diversified, low-fee portfolio.26
Author's note: Advisers may perhaps be regarded either as salespeople or as professionals. If they are salespeople, then this research finding makes perfect sense. If the adviser is a professional, then the professional encountered a client who was employing deceptive tactics. Professional advice is a highly personal activity, and a deceptive client makes it very difficult to provide professional advice. It would have been interesting if any of the advisers had seen through the deception, but this study, unlike ASIC's shadow-shopping survey, in which genuine clients who were actually seeking an adviser provided feedback to ASIC, fails in its design to achieve anything other than identifying salespeople.
A study of 617,375 accounts in US 401(k) schemes, with a single administrator, compared members with delegated and self-directed investment.27 A self-directed member selects their own investments from the plan's range. A member with delegated investment entrusts their entire portfolio either to a target-date fund, which maintains a standardised mix for everyone retiring in a given year, or to a bespoke managed account. In the first quarter of 2020, 2.1% of participants in target-date funds made changes to their portfolios. Among participants who had not delegated management, 16.6% did so.27 Vanguard reported that fewer than 2% of its target-date investors carried out transactions, a rate which, according to Vanguard, is five times lower than that of its other investors, corresponding to roughly one in ten of them.28 Among all members of the pension schemes, 5.3% carried out transactions between January and April 2020.28 Both authors attribute the difference to the structure of the pension schemes. One author of the first study pointed out that they work for the company whose services the scheme administrator uses.27
Australian share prices fell by 30% in March 2020, the sharpest one-month decline in the history of the Australian share market since 1987. The Reserve Bank of Australia examined the behaviour of superannuation fund members.29 Around half of the increase in funds' cash holdings in the March quarter came from members who shifted their investments into cash, equivalent to around 1.5% of total assets under management.29 In data from 30 funds, switching reached 3–4% of assets under management for several large funds and 8% for one medium-sized fund.29 The flows were greater than during previous periods of market turbulence, including the global financial crisis. A small group of active members, who were generally approaching retirement and held larger balances, drove the switching.29
APRA's own account covers members switching back out of cash in the June quarter of 2020.30 So-called platform accounts are excluded from this. Neither of the two regulatory authorities records whether a member had an adviser.
Bodie, Treussard and Willen write that "for most households, human wealth dwarfs financial wealth".31 Their table shows the future earnings of a male university graduate as 47.4 times his current income at the age of 25 and 25.9 times at the age of 35.31 In a study published in 2001, Venti and Wise ranked 3,992 American households aged between 51 and 61 according to their lifetime earnings.32 They divided the households into ten equal groups, known as deciles, from the lowest decile of lifetime earners to the highest. Each household within a decile earned roughly the same amount over the course of its working life. Within the sixth decile, that is, the households just above the midpoint, the median household had accumulated $144,188, while the lowest tenth of that same decile had accumulated less than $30,000. In the ninth decile, the median household had $305,536, while the lowest tenth had less than $100,000.32 The amounts are given in 1992 dollars, as reported by the households. In 2025 dollars, they are approximately 2.3 times higher: $330,000 and $69,000, then $700,000 and $229,000.33 Households that had earned roughly the same amount over the course of their working lives found themselves in very different situations upon retirement. The authors attribute the majority of these differences to saving behaviour, rather than poor health or inheritances.32
Experimental estimates put Australia's human capital stock in 2001 at around $7.0 trillion.34 Household net wealth stood at $2,666.1 billion in June 2001.35 The former is 2.6 times higher than the latter, but the two total figures are based on different definitions.34,35 No Australian time series compares a household's future income by age with its financial assets.
The Retirement Income Review reports that more than half of retirees over 65 deplete their savings at the minimum withdrawal rate.36 It states that minimum withdrawal rates reinforce this behaviour and strengthen the tendency to preserve savings. In inflation-adjusted figures, households' net financial assets, excluding their own home, grow during retirement.36 Age pensioners' assessable assets show no significant change in the five years prior to death. Only around 10% of single age pensioners spent 90% of their assets over an eight-year period.36 Most people with low savings withdraw more than the minimum amount.36
In three experiments, 1,603 Australian members aged 55 to 67 were asked to plan withdrawals from a hypothetical sum of A$350,000.37 When shown the Age Pension of A$24,550 per year, the participants chose an average of A$25,000 in the first step. When shown the ASFA comfortable budget of A$45,239, they chose A$32,500.37 In the first two experiments, each figure was labelled with its source: the Age Pension or the ASFA budget. In the third experiment, the same dollar amounts were shown without any indication of their source, and the decisions followed the same pattern. It was therefore the dollar amounts, and not their origin, that influenced the decisions. The decisions are hypothetical, and well-being was not measured.37
The author of these Tools, Dr Christoph Schnelle, conducted two systematic reviews on the influence of doctors on patients' physical health, taking into account all known factors such as demographic characteristics and patient risk, in his doctoral thesis: one on surgeons and one on other doctors.3,4 This involved measuring the proportion of variation in patients' physical treatment outcomes attributable to the individual doctor. The statistical measure used to quantify this effect was the intraclass correlation coefficient (ICC).38
The review of doctors other than surgeons included 36,239 doctors and found a proportion ranging from 0 to 33%, with a median of 1.9% and a mean of 3.9%.4 The review of surgeons included at least 52,436 surgeons and found a proportion ranging from 0 to 47%, with a median of 4.0%.3 Sixteen of its studies reported on surgeons whose outcomes were entirely above or entirely below the average.3 Both reviews describe their findings as highly variable, meaning the variation of the doctors' effect between surgical and other specialities, different interventions (treatments) and different measurements from death to reoperation to patient blood test values is large.
One well-conducted study included in the reviews examined two professional groups simultaneously: 110,769 cardiac operations at ten centres in the United Kingdom, involving 127 surgeons and 190 anaesthetists.39 The patient's own risk accounted for 95.75% of the variation in in-hospital deaths. Surgeons accounted for 4.00% and anaesthetists for 0.25%.39
The methodological review from the same study contains recommendations on how a practitioner's share could be measured and reported where grouped records exist.38 This review could serve as a starting point for a similar project in other sectors, such as financial advice.
These claims arise in discussions about advice. The table summarises what the available data shows on each point. They are addressed in the Discussion section.
| Claim | What the record shows |
|---|---|
| Investment advice yields around 3% per annum or 5.5%. | Both are estimates by fund managers for advisers' use, from Vanguard and Russell Investments, not measurements of what investment advice actually achieves. Advisory firms' websites quote these figures: one cites 5.2% per annum, another 5.6%. See Estimates published or commissioned by the industry.1,2,10,17 |
| In the long term, households receiving advice have several times the wealth of other households. | One commissioned Canadian study, comparing projected rather than actual households receiving advice. See Estimates published or commissioned by the industry.13 |
| Regular reviews increase households' savings or contributions. | No study was found on the frequency of reviews. The closest matched study found that advice was not associated with self-reported retirement savings.40 |
| More frequent reviews lead to longer customer retention. | No study was found. |
| Advice increases superannuation contributions in Australia. | No study was found. ASIC Reports 639 and 781, which are often cited nearby, examine other issues.41,42 |
| Advice led to more prudent behaviour during the market downturns in 2020 or 2022. | The comparisons measured relate to delegated and self-directed accounts, which is attributed to plan design, while the advisory status is unknown. See Behaviour in market falls.27,28 |
| One study compared fund movements between advised and unadvised households in 2022. | No such study was found. Comments on this year report on flows between categories of fund, not households. |
| An Australian platform study compared the behaviour of advised and unadvised investors during a market downturn. | No such study was found, neither for the year 2020 nor for any other market downturn. |
| Retirees die with 90% of their superannuation. | Two figures for two population groups who held assets, neither of which is stated in dollars for a specific year. Of those aged 60 and over who died between 2014 and 2018, 80% held no superannuation.36,43 The first figure does not take inflation into account in its calculation. Therefore, once inflation is factored in, the 90% figure could amount to about 40% to 55% of the original sum in buying power, at 2.5% to 3% inflation over 20 to 30 years.53 The reason for the latter figure is likely that many people in this group of deceased individuals never had any meaningful super to begin with. |
| Financial advice broadens the options a household considers. | No study measures this. The available evidence shows limited individual customisation and steering to higher-fee funds.21,26 |
| The measured value lies in comprehensive planning: tax, structure, insurance, estate planning and Age Pension. | The studies merely examine holdings and fees, so they cannot capture the planning aspect. No study measures this.21 |
| One study on Australian advisory services, or on fee-based advice anywhere, has separated the effects of advice from the differences between those who seek advice and those who do not. | No such study was found. Virtually all studies that control for selection bias relate to commission-based or banking channels in other countries.6 |
What the findings show together. The published literature demonstrates that advisers differ from one another in terms of the investments held by their clients and the associated costs. It establishes little beyond that. Where individual advisers can be tracked, as in the Canadian, Swiss and American data, the measured effects on risk-adjusted returns are negative or indistinguishable from zero. If the advice is measured on the basis of subsequent wealth, the correlation is only short-lived or disappears as soon as individuals who have received a lump sum are excluded.
There are no studies that distinguish between "advisers" who are merely product sellers and professional advisers; between advisers employed by a product provider such as a bank, an insurance company or a superannuation fund manager, and advisers who are free to choose which investments and insurance policies they recommend; or even between advisers who choose index funds such as Vanguard and those who recommend more expensive branded funds such as Platinum or Magellan.
Why the wealth studies cannot settle the question. Both wealth panels are observational studies, and observational studies can only demonstrate a causal link in rare cases, and not in these instances. A link may disappear if another cause is identified. One such cause is that wealthier people are more likely to engage an adviser. If this is the case, the causality runs in the opposite direction: wealth leads to the engagement of advisers. The American groups showed a 2.5-fold difference in wealth prior to any follow-up.7 In the British panel, the correlation did not differ significantly from zero once individuals who had received lump-sum payments, such as inheritances, were excluded.5 The author of the Texas Tech University doctoral thesis takes the same view: clients of advisers may differ by nature from people who do not seek advice.7
The industry estimates beside the studies. The figures from Vanguard and Russell are made up of individual components. Each component represents modelled savings or gains, such as a lower-cost fund, a better asset allocation, tax savings or an avoided mistake. Taken together, they present an incomplete picture of what advice might achieve, not what advisers were measured doing. Aspects such as the benefits of sharing life experiences or the opportunity to explore a client's career path are not taken into account.
None of the analyses compare clients who receive advice with similar clients who do not. Vanguard's own earlier caveats, that the gain is irregular and varies by client, describe that difference.8 Russell's behavioural component is calculated using fund flows, as are the differences in investor returns on the Active and passive page. This page explains why such a gap changes with dollar-weighting and fund costs, and highlights the benefits of sticking with investments whose returns are as close as possible to market indices such as the S&P 500 or the ASX 200, rather than switching funds in response to their past performance, which is often a loss-making strategy. The Gamma Factor compares predicted rather than observed advised households in surveys commissioned by a financial firm.13
The estimates are published by fund managers for use by advisers. Russell's report states that it must not be shared with clients.2 Nevertheless, the figures do reach clients, sometimes without the appropriate caveats, as demonstrated by Vanguard Australia's consumer website.9 The studies which directly examined advisers, in What individual advisers do, revealed smaller or negative effects. A client who hears "3%" or "5.5%" is hearing a description of what good advice might achieve. It is not a measure of what advisers actually achieve.
The twelve claims in the previous section just above. Most of the claims are plausible. None of them has been tested in the form in which they are presented. In the author's experience, some of them are promoted by large firms in their marketing to advisers and the public. Three of the claims deserve particular attention.
The first is that advisory services reassured clients in the 2020 fall. The only measured comparisons relate to delegated and self-directed accounts in employer plans, and both publishers attribute this to the structure of the plans, not any adviser relationship, and neither records whether such a relationship existed.27,28
The second claim is that retirees die with 90% of their superannuation. The two underlying figures come from an unnamed fund and from age pensioners, and almost certainly compare dollar amounts without taking inflation into account.36 There are usually twenty or thirty years between the account balance at the time of retirement and the account balance at the time of death. Hence the "90%" could be only about 40% to 55% of the original sum in buying power.
Most Australians aged 60 and over who died between 2014 and 2018 had no superannuation savings at all in the up to four years prior to their death.43 Mandatory superannuation was not introduced until the 1980s, meaning that many of them had pensions or assets outside the superannuation system.
The third claim is that the value of advice lies in comprehensive planning. The authors of the Canadian study themselves mention this possibility: advice on savings for education and retirement, tax and estate planning.21 However, the data available to them and others merely records assets and fees, meaning no study has measured comparative outcomes. The claim is unproven and remains unresolved in either direction.
What advisers claim, beside what has been measured. The websites study quantifies the gap between the two. Almost every adviser's website claims that the adviser is good, and three out of four promise an outcome. One in twenty claims better returns, and fewer than one in six of those cite a source.17 Published research, meanwhile, has directly measured advisers across three datasets and found that they differ from one another, without outperforming the alternatives. The claims are not untruths: a claim about experience or diligence may well be true, and nothing on a website could substantiate this one way or the other. Yet a client reading advisers' websites meets confidence everywhere and evidence almost nowhere. The same applies at a higher level. Certification bodies describe their holders as the "gold standard", and the only body that provides evidence offers a survey that does not reveal how the respondents were recruited.17
The single gap. No study has measured how much of the difference between clients is attributable to the individual adviser, the way medicine has measured the doctor's share, which, despite millions of research papers already having been published, was only examined in detail in 2019. There is no dataset that distinguishes between advised and unadvised behaviour in the event of a market crash, neither in Australia nor elsewhere. No Australian study, nor any study on fee-based advice that the author could find, separates the effect of advice from the characteristics of the people who seek it.6
What medicine suggests. Medical reviews identify outcomes where the practitioner accounts for nothing, and outcomes where the clinician accounts for almost half the variation in patients' physical health outcomes.3,4 The proportion depends on the treatment (procedure) and the outcome being measured. In the cited cardiac study, the anaesthetist's contribution to deaths was negligibly small.39 In the author's view, a metric more relevant to the anaesthetist's actual work, such as post-operative pain, might reveal a much greater contribution by the anaesthetist to patient outcomes. The situation could be similar in the context of advisory services. There may well be a meaningful and available metric, but one that has not yet been identified.
The risks of measuring advisers. The study of surgeons conducted by the author revealed that even a small number of high-risk patients can skew the results of an individual surgeon. It quoted that raw performance data should not be published.3 The same would apply to advisers whose clients are older and for whom capital preservation is often the primary concern, meaning that better results may simply reflect the demographics of the clients: older clients may be content with returns that younger clients would find unacceptable in good times, so a good adviser would deliver different returns to different clients, beyond any adjustment for risk profile.
Thoughts on the value of advice, by the author. This page shows that there is little published, or perhaps even known, evidence on advisers making a difference, and some of the evidence shows that advisers financially harm clients, though that same example, under "Allocations, and what advisers held themselves" above, showed that the advisers make the same mistakes with their own investments.
One conclusion from the above could be that the 'advisers' are really salespeople who work under instructions without realising the flaws in those instructions. The culprits in that case are those who manage and instruct these 'advisers' and these 'advisers' are not working at a professional level or anywhere near it.
This is one consequence of a licensing regime where an organisation, the licensee, licenses advisers rather than like in law, accounting, medicine, even trades people, where the individual is licensed and therefore has a strong incentive to behave in a way that preserves their license. Theoretically that is currently the case but in many cases today it is the licensee that is shut down for rogue behaviour but the majority of 'advisers' move on, though that is getting harder for those advisors and the advisers, with the current degree requirements also have a higher stake in their career.
This was different up to 2019 where there were courses that qualified advisers in as little as four days and advisers who got banned simply moved into other areas of endeavour, having had little to nothing at stake in their career.
Speaking purely from personal experience, I have been an adviser for 20 years and I have observed many wealth transformations for the better among my clients. The simple act of checking in at least once a year and having an expert available at any time seems to have a transformative effect over years. Being able to save on tax, optimise their returns and averting losses by not following ideas that fail under closer, expert scrutiny are cumulatively also very beneficial.
There are many different types of advisers and of advice. A hierarchy of advice could be written this way, ranging from worst to best:
Differences in quality among advisers can, in my experience, be very large, and research that does not take account of this variation, treating advisers as being more or less the same, such research is unlikely to lead to a meaningful result. The first research question should therefore be what makes a good or exceptional adviser, looking for objective criteria of what defines the differences among advisers. Then the effect of advice can be measured with practitioner quality taken into account.
In the author's opinion, based on his PhD research on what makes an exceptionally good medical doctor, a good adviser needs to be good with people and able to understand people, the deeper the better, so communication and understanding skills are critical. An adviser needs to have a good heart and deep respect for clients and that they have the unfettered right to make their choices and therefore offers but does not impose. Advisers who impose behavioural finance concepts on their clients without respect for why clients made their current choices can be very successful but may not provide the most harmonious advice.
A good adviser also needs substantial technical knowledge and skills. Experience, whether advice experience or life experience, can make a very big difference, as lived experience provides an authority that knowledge alone does not have.
There are very large differences in quality between individual lawyers, accountants and doctors. It is the same with financial advisers. Dealing with a bad professional can be very harmful. Meeting an average professional is useful. Meeting a brilliant professional is a joy and can be transformative.
In the author's view, the first and third questions offer starting points for research at doctoral level (PhD).
A clear indication that advisers could play a role comes from a study of Norwegian GPs, the only study to examine the extent to which an individual doctor influences their patients' life expectancy that can establish causation.44 Under Norwegian law, patients are assigned a new GP via a computerised process if their own GP closes their practice or reduces their patient list; the patient therefore does not choose the GP themselves.
Ginja, Riise, Willage and Willen followed 92,205 patients aged 55 and over who were moved to 2,023 new GPs between 2006 and 2015, and assessed each GP based on whether the patients assigned to them were still alive two years later. A patient assigned to a GP who was one standard deviation higher on this measure died 4.1 months later (interval 1.5 to 6.8) or 8.3 months later (4.8 to 11.8) if they were aged between 55 and 61 at the move. Across the entire range of GPs, from the 10th to the 90th percentile, the difference was 7.0 months (2.5 to 11.5) and 14 months (8.2 to 19.9) for patients aged between 55 and 61; the two-year risk of death was 2.0% instead of 4.4%, an interval that reaches zero. The benefit was evident in cancer-related deaths, which the authors attribute to earlier detection, and the better GPs did not incur any additional costs for the system. The authors estimate the value of replacing the worst 5% of GPs with average GPs at $1.2 billion or $12.4 billion, depending on whether a life year is valued at US$35,000, as set by the Norwegian government, or the US$369,000 set by the US FDA.
The age, gender, specialism and practice style of the GPs accounted for around 1% of the difference; there were no measurable differences between male and female GPs. Patients did not seek out the better GPs when given a free choice, and the online reviews were not correlated with the quality of the GP. The authors offer two readings: patients either cannot identify high-quality GPs, or they value other aspects of a GP than its ability to save lives. In the author's view, it is the first. The full entry, including all intervals and limitations, can be found on the practitioner parallels page.
Perhaps a financial GP (a professional adviser) could have a similarly strong effect, not in preserving life but in other areas, and could possibly also mitigate the costs of engaging their services through their advice.
What the spread in other lines of work suggests. Wherever professionals have been assessed on the basis of clients they did not select themselves, there is considerable variation between them, and the best are just as far above the mean as the worst are below it. A manager in the 90th percentile boosts a team's performance by around 24 transactions per hour, double the 12 contributed by a manager at the 10th percentile, while an employee handles an average of around 10 transactions per hour.47
A court-appointed defence lawyer in the 90th percentile secures a sentence that is 5.8 months shorter than one in the 10th percentile, a difference equivalent to 82% of the average sentence of seven months.48
The top quarter of bariatric surgeons had only one-third the complication rate of the bottom quarter, and blinded peers, each watching a video, were able to identify them.49
A teacher who is one standard deviation above average is worth over $400,000 per year in their pupils' lifetime earnings.50
Of 198 samples of performers, 94% followed a heavy-tailed distribution, in which the top 5% accounted for around a quarter of the performance.51 In each of these studies, qualifications and experience accounted for only a small proportion of the differences. No study on advisers employs this design. The study that comes closest to this, the Canadian study mentioned above, found that the adviser explains more of a client's portfolio than the client's own characteristics.21
A 2024 market overview of the advisory sector lists the distribution of adviser quality among its open questions.52
If advisers are distributed in the same way as any other measured profession, this has four consequences. The question "Does advice add value?" is the wrong question, because the answer depends on which adviser is being considered. The best advisers are worth a great deal, and no one has quantified their value; the "around 3%" cited by the industry is an average figure with no variation. The best cannot be identified on the basis of what advisers say about themselves, which is almost all a client can see. And the evidence base of advice, like medicine's, maps the worst and never the best.
How much of the difference between clients belongs to the individual adviser? The published research measures the average adviser. Medicine shows that the individual practitioner's share can be measured and that it varies depending on the outcome (see The individual practitioner's share, measured in medicine). The equivalent for financial advice would require studies that group clients by adviser and measure a household's financial outcome, not the contents of a portfolio.
What would a household panel have to ask? Panels record whether advice was provided within a specified period, so an initial appointment and a twenty-year relationship give the same answer. Three additional questions would be helpful: whether an ongoing relationship exists and how long it has lasted, what strategy was followed, and whether a financial windfall occurred. Furthermore: what was paid for the advice? The Financial Planning Longitudinal Study in the United States, which has been running for a decade, records the type of adviser, not the specific adviser.45
Could the Australian version be built from records that already exist? There is a public register of advisers: ASIC's Financial Advisers Register.46 The client data is held by the platform operators. This raises three initial questions: Would the operators make the data available grouped by adviser? Under what data protection conditions? And what outcome would be measured, given that a platform dataset reflects holdings and transactions, but not the living circumstances of a household? Such a study would compare advisers with one another, which is a different research question from comparing households receiving advice with those not receiving advice. Using the same data sets, platform clients receiving advice and those not receiving advice could be compared with the Reserve Bank's figures through March 2020. In the author's experience, platform accounts are the ones most frequently used by advisers. "Wrap" providers used in Australia include Macquarie Bank, BT Panorama, Netwealth, Colonial First State and AMP North. Industry funds could carry out the same comparison for members recorded as receiving advice.
Advisers differ from one another in terms of their clients' assets and the associated costs. Whether advice leaves households permanently wealthier has not been established; the industry's figures estimate what advice could add rather than measure what it does. The extent to which a client's outcome can be attributed to an individual adviser has not been measured. Medicine shows that it can be shown. Research for measurable quality criteria for advisers could be useful. It might also be worthwhile to examine whether publicly available information about advisers could be used to assess their quality, at least to a limited extent.
Designed and produced by Dr Christoph Schnelle, In Your Interest Financial Planning. In Your Interest Financial Planning Pty Ltd, ABN 28 094 300 464 is Authorised Rep. No 308161 of Fiduciary Duty Advisers Pty Ltd AFSL No 527434. Nothing on this page is advice to you or to your clients, and nothing on it is a recommendation of any product, any strategy or any adviser. Corrections are welcome and wanted – contact us.