There are a growing number of businesses offering financial modelling as part of their services, particularly in the property industry i.e. buyer’s agents.
It is not uncommon for buyers’ agents to provide you with a model of how you might build wealth through investing in multiple properties. These projections typically include cash flow and net worth over time, which can make it easier to visualise how the strategy could play out.
The problem is that modelling can be made to produce almost any result. The output is not the truth. It is simply the consequence of the assumptions.
Please learn from the financial planning industry’s mistakes
One of the financial planning industry’s long-running failures was dressing up product sales as “advice”. That era largely ended when commissions on financial products were banned in 2013, which, in my view, is one of the best decisions the Australian federal government has ever made. This is probably why the ETF industry has grown from $7 billion to over $330 billion since 2013.
Before 2013, it was common for investors to seek out a financial adviser expecting strategic guidance, only to be sold a product that paid the adviser a commission. That mismatch is how trust is misused – clients think they are paying for honest strategic advice, but they are being sold a commissioned product, and the damage only becomes obvious after the investment disappoints.
Everyone understands what happens when you walk into a used car yard: the salesperson is there to sell you a car. It is buyer beware. Financial advice should not work like that.
If someone needs advice and believes they are getting advice, then advice is what they should receive, not a sales pitch dressed up in advice language.
Assumptions drive modelling outcomes
The most important thing to understand about any financial model is this: assumptions drive outcomes. If you want a particular result, you can usually get it simply by changing the assumptions until the numbers behave.
Now add incentives. If the modeller gets paid when you transact repeatedly, the model is naturally at risk of being structurally biased toward optimism rather than conservatism.
There is a second conflict too: if the modeller is also the person choosing the asset that will “make” the strategy, there is obvious pressure to dial up return assumptions, so the strategy looks compelling.
Put differently, for instance, if a buyer’s agent models an investment property purchase using only 5% p.a. capital growth rate, most people would immediately ask: why am I paying you a fee if your expected growth rate is below the long-term median? That question alone tells you how quickly the conversation becomes about selling a decision, not testing a decision.
Financial modelling is a very useful exercise if it is based on high quality assumptions. High quality assumptions are conservative, but not so conservative they become meaningless – there’s no point assuming an average 10% p.a. mortgage borrowing rate if we don’t think that is likely. In practice, assumptions should sit toward the more conservative end of a defensible range. They should be evidence-based: consistent with what has been observed over long multi-decade periods across Australia as a whole, not cherry-picked from a hot postcode, a particular product type, or a narrow sub-asset class.
One more test: if the proposed strategy models only one asset class or a single property strategy, and does not properly model credible alternatives, then it is likely to be closer to a sales pitch than investment advice.
Returns are not straight line – sequence risk matters
Most investors understand returns almost never arrive in a neat straight line. Short-term performance is largely random and can be volatile. But if you are building a model that runs for 10, 20, or 30 years, the only workable approach is to use long-term average assumptions, because it is almost impossible to forecast things like interest rates, credit policy, rents, or prices year by year with any reliability.
That is the core tension. We know returns will be uneven. Yet we are forced to model them as if they are smooth, simply to make long-term modelling possible.
And this is where sequence of returns matters. What if you (accidentally) buy at the end of a growth cycle and get very little capital growth in the first decade? The long-term average might eventually be acceptable, but the lived experience of the strategy can be completely different.
If a strategy relies on early returns to fund the next steps, that is a red flag. For example, if I buy a property today assuming I can access borrowable equity in two years to help purchase a second property, I must confront the risk that the growth simply does not materialise on schedule.
Be wary of any “strategy” that depends on strong short-term returns to make the plan work. If the early growth does not show up, the whole thing can stall.
The cleanest way to manage sequencing risk is to be more conservative with long-term modelling assumptions, especially for the early years when the strategy is most fragile.
It’s easy to overestimate future property cash flow
When modelling investment properties, it is easy to overestimate future cash flow, and the risk increases the longer the projection period. A few years ago, I wrote about how sharply the holding costs of property have risen. In response, we now typically allow around 30% to 35% of gross rental income for expenses.
Even then, forecasting is messy because many costs can move quickly and unpredictably. Land tax rules can change, insurance premiums can jump significantly, council rates can rise materially from one year to the next, and property maintenance is one of the hardest items to budget with any precision.
One of the biggest assumptions driving long-term cash flow is rental income growth – how a property’s gross income will change over time. Many models effectively assume the starting yield remains constant over time. But if capital growth is strong, yields usually compress, unless rent growth keeps pace, as tenants’ ability to pay becomes the limiting factor. That is why rental growth assumptions should be anchored to the income growth of the likely tenant demographic, not the assumed capital growth.
In broad terms, Australian income growth tends be around 2.5% to 3.5% p.a. So, unless there is clear evidence to justify something higher, a sensible base-case rental growth assumption for investment-grade property is typically around 3% to 4% p.a.
In lower socioeconomic locations, it can be more realistic to adopt a lower assumption again, because rent growth is more likely to be constrained by tenants’ capacity to pay.
Assess execution risk
When you develop a strategy, you must consider execution risk.
Execution risk is the risk that you cannot implement the strategy because a key constraint changes. In property investing, the most common constraint is borrowing capacity. Borrowing capacity can shift because interest rates move, lenders credit policy changes (interest rate buffers, living expense scrutiny, shading of rent, DTI caps, treatment of negative gearing), or your personal circumstances change.
If the strategy is already tight on borrowing capacity, or it relies on enhanced borrowing capacity to hold true, execution risk is materially higher.
We have seen this risk play out over the last decade. For instance, interest-only lending has been restricted, so rolling over interest-only terms is not as straightforward as it once was. Previously, many investors assumed they could keep their investment loans interest-only indefinitely. That assumption turned out to be wrong, and it is a clear example of how execution risk can derail an otherwise “sound on paper” strategy.
This is why you must stress-test the strategy for common execution failure points. What if your borrowing capacity does not allow you to buy the next investment property? What if you cannot refinance or roll over an interest-only loan and you are forced into principal and interest repayments?
If either of those events occurs, does the strategy still work, or does it break? And if it breaks, what is your Plan B?
Perform sensitivity and scenario analysis
A sensitivity analysis involves changing key financial modelling assumptions (interest rates, investment returns, tax rates, and so on) to identify which variables are most sensitive, meaning they drive the biggest change in outcomes.
It is important to run this analysis because it tells you what really matters. Once you know which assumptions have the greatest impact on success, you can focus your effort on reducing the risk around those variables.
For example, if the strategy is highly sensitive to capital growth, then asset selection is absolutely critical. In that case, it makes sense to seek guidance from a reputable buyer’s agent, because a small difference in capital growth can make a material difference to the end result.
If the strategy is highly sensitive to borrowing capacity, then changes in credit policy or interest rates can directly affect your ability to implement the strategy as modelled. In that scenario, you may need to be more proactive in managing borrowing capacity, such as locking in access to equity as early and often as possible and bringing forward the acquisition phase while borrowing capacity allows.
If outcomes are highly sensitive to interest rates, then building a sizeable cash buffer becomes a priority. A buffer gives you options and reduces the risk of being forced into reactive decisions when rates rise.
If the strategy is highly sensitive to surplus cash flow, then strong cash flow discipline matters. That may mean implementing a tight cash flow management system, monitoring actual versus projected cash flow, and identifying expenses or structuring changes early before the strategy starts to drift off track.
At best, a financial model is an informed estimate, not a definitive answer
Financial projections can be an important facet because it translates a strategy into a best-estimate picture of what could happen if you follow it. But it is still not a verdict. It is the logical output of assumptions that underpin it, and the assumptions are where the real judgement (and possible bias) lives.
If someone gives you a model and they have a vested interest in you acting on it, you must interrogate it, not accept it at face value. Are the assumptions conservative and evidence-based? Does it survive stress tests for interest rates, borrowing capacity, and timing risk? Does it compare credible alternatives?
Do that, and modelling becomes a helpful decision tool. Skip it, and you risk buying an outcome engineered on paper, not rooted in reality.
