Educated guesses are key to financial planning.
No matter how diligent we are when gathering information about a client, financial planning is about the future, not the past. As such we will need to make some “guesses, or assumptions, about what will happen in the future. Whilst these guesses need to be as educated as possible, they are guesses all the same.
While historic data can be useful, financial plans look forward. The warning that ‘past performance is not necessarily a reliable guide to the future’ is particularly relevant. The evidence does not suggest anyone has managed to accurately predict the sort of variables involved in financial planning. Instead, a robust process gives us a base upon which we can start working with our clients.
Until relatively recently, the only real option to measure UK price inflation was the Retail Prices Index (RPI). This measures the price of a ‘basket’ of goods and services over time. It has the advantage of a long history, with official monthly data being available back to January 1956. Annual data goes back to at least 1900.
An assumption for the change in the RPI is helpful as most other variables revolve around it. Economists in both the public and private sector devote considerable effort to guessing what it might do. One method is to look at the spread between an index-linked and conventional gilts. Another is to use the long-term historical average (4.1% from 1900 to 2019). A simple approach is to take the 2.5% RPI target for the Bank of England’s as a starting point. We do so on the basis that the Bank will, over the long term, achieve that target.
The Consumer Price Index (CPI) has been used to measure inflation since the late nineties. It became the principal measure of inflation in 2003. It excludes mortgage interest payments and other housing-related costs, such as council tax. However, it is being used increasingly as the preferred measure for escalating of tax bands, state benefits and certain pensions in payment. This means that an explicit it is an important variable in many financial plans. To date, the CPI rate has tended to be lower than RPI. The Bank of England’s long term target for it is 2%pa.
Relationship with expenditure
To avoid skewing the data, the buying patterns of certain consumers do not factor in when it is calculated. These “outliers” are made up of very high and very low earning households. High earners are defined as the top 4% of households by income. Low earners are pensioner households where 75% of their income is derived from state pensions and benefits.
The exclusion of the wealthiest 4% of households may be an important factor in many financial plans. According to a 2012 article in The Guardian, the starting point for high earning households was around £50,000 after tax. The evidence is that the expenditure of such individuals is more heavily weighted towards services. The prices of these tend to increase at a faster rate than inflation and goods as a whole. They may include domestic help, childcare, the costs of a second home or comprehensive private health insurance or care. We will make a separate assumption for how much each of these types of expenditure increases by.
Average earnings data is collected by the Office of National Statistics (ONS). The historic series goes back to 1963 in the form of the National Average Earnings Index (NAEI). The current measure (since 2010) is Average Weekly Earnings.
This reveals earnings have generally, although not always, increased at a faster rate than prices on average.
We should be aware that what may be relevant for the average employee may not relevant for a specific client. For example, someone just entering a good career may expect double-digit increases over the short term. Conversely, someone coming towards the end of their career may not expect much of an increase to their pay. Our standard assumption is that at the very least earnings match inflation. However, a discussion is needed to ascertain whether a different rate is more suitable.
Education costs can vary significantly. School fees, for example, tend to increase in jumps at various stages.
A large proportion of education costs come from paying staff. Therefore, any rise may be linked to the rise in earnings unless there are other factors to consider suggesting otherwise. There is also the potential for ‘extras’ such as trips to increase costs significantly. While data is available from surveys that indicate the average fee rises across the institutions who respond. This currently stands at over 5.2% pa. However, there is no guarantee the particular institution will match the average.
In recent years there have also been some major changes to the funding of undergraduate studies. The Government has replaced grants with loans. More recently, they introduced tuition fees after many years of tuition costs being met from general taxation. Since their introduction, tuition fees have remained largely static. An assumption for these can usefully be made, even if the vagaries of politics mean that it is liable to change without much warning.
As well as the cost of the course itself, the student still needs to live, and many parents will wish to contribute to this. Our standard assumption is for these costs to rise in line with CPI. We use a figure of £20,000 pa if the client does not have one of their own.
The consensus seems to be that there is a one in three chance of requiring care at some point. We tend to err on the side of caution when building plans so we always assume someone will go into care. We use a calculator to predict the cost of a care home and use the most expensive figure. Studies suggest an average stay is in the region of two and a half years but we use three.
As with education costs, the largest component of the cost is wages. As such we will tie in increases to an earnings inflation assumption of RPI plus 1%.
According to the ONS, a 65-year-old male currently has a one in four chance of living until age 94. However, these are averages. As we are being conservative, we always plan to age 100 (or over). We also consult with actuarial figures for how health can affect this. This becomes a discussion point when looking at the results of any predictions.
State benefits may well be relatively insignificant compared to overall resources. Nevertheless, most clients will be entitled to a state pension. They may also be entitled to such benefits as employment and support allowance in the event of incapacity. Fortunately, the inflation rates at which these increases are set by the government. Our task is to assess eligibility.
The Coalition Government introduced the ‘triple lock’ inflation protection for the basic state pension in 2010. However, politicians have a generally unimpressive record when it comes to long-term consistency. There is no certainty that this will continue in its current form for the duration of a typical financial plan. We tend to tie State Pension increases with inflation rather than any higher increases such as average earnings.
We utilise the services of Distribution Technology one of the country’s leading sources of data to construct asset allocations. They produce return expectations for all the major asset classes. For our portfolios, the assumed returns range from 3.51% to 9.41% pa before charges. They have stated the following about their methodology.
For sovereign bonds, the expected returns are derived from their gross redemption yields based on a representative index which has a stable duration.
Market yields for Index-linked gilts (“ILGs”) are quoted based on inflation assumptions of 0% and 5% and our ILG assumption is interpolated to create the 2.5% inflation expectation.
In the case of sterling investment grade and global high yield bonds, the prevailing option-adjusted yield spread (OAS) for the relevant credit index (or corresponding credit portfolio) is added to our UK gilt assumption.
The methodology for generating expected returns incorporates the earnings yield via the dividend payout ratio, as well as considering the regional dividend yields, GDP and inflation forecasts. Our research indicates that this method is more comprehensive as it captures any share buybacks over time as well as the actual payment of dividends. We have used consensus data, sourced from Bloomberg, for regional GDP and inflation forecasts. These consensus forecasts include contributions from various Investment Banks, consultancies and other local agencies.
Cash (Deposits) represents cash held in savings accounts of retail investors and the assumption for this is for marginal returns and volatility.
Any cash held in investment holdings is assumed to be Cash (via Money Market instruments).
Our assumptions are based on a basket comprising 50% UK Commercial Property and 50% global REITs.
The assumption for commodities is in line with the global growth forecast from the IMF as research indicates that commodity prices are highly correlated with global growth.
The assumption for this specialist set of strategies is based on the observed historic data of a broad multi-strategy hedge fund index.
Variability of returns
We also get data from distribution technology on how these returns can vary. We use this information to run scenarios involving varying returns and gauge their impact.
Another thing we do is to include a scenario every 10 years where there is a market downturn. These are akin to the credit crunch of 2008/9. The idea of this is that over the long-term returns will stay in line with the central assumption. Over the short term, the portfolio falls in value by between 10 to 40% depending on its equity exposure. The portfolio has a period of recovery and high growth for three years. We do this to model the impact of these times of volatility on the sustainability of withdrawals.
Most investors probably view property is an asset class which has generated high and fairly stable returns.
However, it is important to distinguish for financial planning purposes between several types of property. These can include the family home, rental residential property and commercial property.
The future return on the family home might be of no relevance beyond looking at estate planning. However, there may be an expectation to downsize or upsize in the future. For downsizes, we are loath to include any proceeds into the plan. In our experience, unless someone is moving from an expensive area to a cheap one, the proceeds are negligible.
Historic price data is available from sources such as Halifax, Nationwide and the Land Registry. However, this only goes back to 1983, 1973 and 1995 respectively. The ONS has a series going back to 1930. This shows a sizeable increase in the value of homes of 6.48% pa between 1930 and 2020. Inflation averaged out at 4.8% over the period. What we can see from the data is that prices only took off in the 1970s. If we take inflation into account real growth on property values have been 2.15% pa since 1970.
There are also some issues with index data for property values, including not incorporating capital expenditure. However, an assumption of RPI plus 2% could be suitable for capital growth on residential property.
For property investors who expect a return in the form of rental yield, does this affect capital growth? There is a logic to deducting the actual or assumed yield from capital growth. This avoids double counting the yield both explicitly and within the total return. Commercial property leases commonly oblige the lessee to maintain it at their own expense. This reduces the owner’s exposure to those costs. With residential lettings, such maintenance is usually the owner’s responsibility, which reduces their return. Even discounting maintenance, there are other costs to consider. These include agents’ and legal fees, and changes to legislation (such as requiring additional fire precautions, wider doorways). There is also the impact of void periods (i.e., when the property is empty) to consider.
Other illiquid assets
For other illiquid assets, such as unquoted businesses, there are as many potential inflation rates as there are assets. Business valuation methods are governed by international accounting standards and guidelines. However, it can ultimately come down to whatever someone is willing to pay. We should consider that the higher risk implies a higher expected return. We should also consider the higher risk might give rise to a total loss. As such we may, without a compelling counterargument, reasonably elect to assume returns keep pace with price inflation.
Finally, individuals will own some assets whose value can be expected to depreciate. These usually comprise modes of transport (of a variety of types, including motor vehicles, boats, and aircraft). A negative growth rate is of 10% pa is potentially suitable. In the case of motor vehicles, where regular replacements may be required, we model periodic expenditures or lease costs. The retention of such an asset until the market value is negligible is unlikely.
Tax rates and bands
Tax rates can inevitably be expected to rise and fall over a client’s lifetime. They do so according to the political complexion of the government of the day and the prevailing economic environment. To attempt to predict anything other than the continuation of the current regime is extremely difficult. We just assume current bands increase by CPI.
Vesting pension benefits
Until the 2014 budget, annuity purchase was the option pursued by most investors in defined contribution pension schemes. Our standard assumption is to assume that all income from pensions will be delivered using drawdown over the clients lifetime. It should be appreciated, however, that this may well be inappropriate for some investors. The individual’s circumstances will dictate the most appropriate route or routes.
One approach is to assume that the individual will take an income on the simplest and most expensive basis. This would be buying an index-linked taxable pension paid annually in advance with 100% survivor’s benefits. If this approach suggests the desired goals are achievable, it provides scope for flexibility. This could include higher expenditure, more gifting and/or a lower risk approach in the non-pension assets. If it does not, there is the opportunity to vary some or all these parameters to rectify things.
The greatest positive impact is generally by opting for a lower rate of escalation. This is because the annuitant generally must live longer than expected to get better returns compared to level annuities.
Spurious precision and the need for reviews
Nobody has yet found any reliable way to predict the future. As such, when we produce any sort of projection, we need to acknowledge that it will be wrong. It does not tell the future; it helps us to project the present forwards.
We do not give you a crystal ball, which magically tells you the future. We give you something we can collaborate on, which we revise and revisit year on year. It is the process of doing this where the value lies.
If you want to know more about this, feel free to book in a free no-obligation chat here or get in touch.
Source of information: IFP Fellows’ briefing paper: Assumptions (https://www.cisi.org/cisiweb2/docs/default-source/SIR/fellows-briefing-paper_assumptions.pdf?sfvrsn=0)