90% of Property Advice Online Is Noise. Here's How I Filter the Signal.

My mum tells me about fifteen things every single day. Close the window. Make your bed. Stop sleeping in. Why haven't you called your grandmother.

None of it changes my behaviour. It's background noise. Completely ignorable.

But when she says "dinner's ready" — I'm at the table in under thirty seconds.

That's the difference between noise and signal. And if you can internalise this distinction for property investment, you'll make better decisions than 90% of the people scrolling through real estate content on social media right now 1.

I come from an actuarial background. We're trained to extract meaningful patterns from oceans of data while ignoring everything else. The property market generates an extraordinary amount of data — and an even more extraordinary amount of garbage disguised as data. Today I want to share the three filters I use personally to separate one from the other.

Let me give you a concrete framework for how I personally consume property information, because "filter the noise" is easy to say and harder to implement in practice.

I have exactly four sources of property market data that I consider signal:

1. CoreLogic Home Value Index (monthly). This is the most comprehensive property price dataset in Australia, covering over 15 million property records. When CoreLogic reports that Melbourne house prices rose 0.8% in October, that's signal. It's based on actual transactions, not surveys or predictions.

2. ABS Building Approvals (monthly). This tells me how much new supply is coming into the market. In Melbourne's southeast, building approvals for detached houses have been declining — which means supply constraints are tightening, which supports price growth.

3. RBA Minutes and Statements (monthly). Not the media interpretation of the minutes — the actual minutes, published on the RBA website. Reading primary sources takes 20 minutes. It saves me from reacting to someone else's selective interpretation.

4. SQM Research Vacancy Rates (weekly). This tells me the rental market supply-demand balance at the suburb level. When vacancy in Cranbourne drops below 1.5%, I know rental demand is outstripping supply — which means rents are rising, which means yields are improving.

Everything else — social media commentary, newspaper opinion pieces, podcast predictions, YouTube property gurus — goes through the three filters before I give it any weight. Most of it gets filtered out. The small amount that survives is usually confirming what the primary data already told me.

This might sound like a lot of work. It isn't. Checking four data sources once a month takes less than an hour. Scrolling through property social media for an hour delivers approximately zero actionable intelligence. The time investment is identical; the information quality is incomparable.

Filter one: has the dust actually settled?

A few months ago, every property blogger, mortgage broker, and self-appointed interest rate expert was screaming about a rate cut. "97% probability!" they said, citing futures markets and economist surveys 2.

Then the Reserve Bank met. And didn't cut.

Every single piece of content produced in the lead-up was noise. Not signal. Because the event hadn't actually happened yet.

My rule is simple: until something is confirmed by the decision-maker — not predicted by commentators — it's speculation. And speculation is noise. The RBA Governor speaks. That's signal. An economist's forecast? Noise.

This applies across the board. "Melbourne prices are about to crash" — has a crash occurred? No? Noise. "This suburb is the next growth corridor" — has the infrastructure been funded? No? Noise 3.

Waiting for confirmation feels slow. But I'd rather be slightly late with accurate information than aggressively early with speculation that turns out wrong.

Let me give you another example that's even more current. Earlier this year, a major bank economist published a forecast suggesting Melbourne property prices would fall 15% by year-end. This prediction was picked up by every property news outlet, shared thousands of times on social media, and cited by dozens of smaller content creators as evidence that the market was about to collapse.

The prediction was made in January. By September, Melbourne house prices had actually risen 4.2%, according to CoreLogic's Home Value Index. The economist quietly revised their forecast in July, but the original doom-and-gloom prediction was still circulating on social media in October.

This is the lifecycle of noise in the property market. A prediction gets made. It gets amplified. It gets stripped of context. And months later, long after it's been disproved by reality, people are still making decisions based on it.

The signal — the actual price movement data published monthly by CoreLogic — told a completely different story. But signals are boring. They don't generate clicks or emotional reactions. So they get buried under noise.

Filter two: follow the claim back to its source

Here's one that had people panicking recently: "Victorian land tax has doubled!"

Sounds terrifying, right? Except that's not what happened. What actually changed was the emergency services levy — from $150 per year to $300 per year. An increase of $150 annually. In the social media echo chamber, that became "land tax has doubled" 4.

This kind of distortion is everywhere in property media. A genuine but minor policy change gets amplified through successive layers of content creators until it's unrecognisable from the original source.

My second filter: when I see a claim that provokes a strong emotional reaction, I trace it back to the primary source. The actual government gazette. The actual APRA determination. The actual ABS data release. Nine times out of ten, the reality is far less dramatic than the headline 5.

This takes five to ten minutes. Almost nobody does it, because scrolling is easier than researching.

I'll add another dimension to this filter that I think is particularly relevant for property investors: be deeply suspicious of any claim that includes the phrase "experts predict" or "according to analysts." These phrases are linguistic camouflage for speculation.

Who are these experts? What's their track record? Do they have skin in the game? A mortgage broker predicting a rate cut benefits from that prediction because it drives more loan applications. A property developer predicting price growth benefits because it drives more sales. An economist predicting a crash benefits because it generates media coverage for their employer.

Every prediction has a motivation behind it. Understanding that motivation is often more useful than evaluating the prediction itself.

The one exception I'll make: predictions that go against the predictor's financial interest. When a property developer warns that a specific market is overheated, pay attention. When a mortgage broker suggests waiting rather than borrowing now, listen carefully. Advice that costs the advisor money is almost always genuine.

Filter three: extend the observation window

If you want to know whether someone's property advice is worth following, don't watch one of their videos. Watch twenty. Go back six months. Go back a year.

I know specific commentators who spent all of last year attacking Melbourne's market. Now that Melbourne's shown signs of recovery, those same people are claiming they "always saw the potential" 6.

Consistency over time is one of the strongest indicators of genuine expertise. Not because experts are always right — nobody is — but because genuine analysts don't flip their fundamental thesis every time the market moves 3% 7.

I've published over 200 pieces of property content since I started. From my very first article to my most recent, the core thesis hasn't changed: Melbourne's southeast offers the best risk-adjusted returns in the Australian residential market. I've been saying this consistently for years, not because I'm stubborn, but because the data hasn't changed.

There's a deeper structural reason why so much property content is noise, and it's worth understanding if you're going to spend any time consuming media about the Australian property market.

Content creators are incentivised by engagement, not accuracy. A video titled "Melbourne property prices are about to CRASH" generates ten times more views than one titled "Melbourne property prices are likely to continue their historical trend of moderate growth." The algorithm rewards alarm. Accuracy is commercially irrelevant.

This creates a systematic bias toward extreme predictions, sensational claims, and emotional manipulation in property media. The content that reaches you through social media algorithms has been selected specifically because it provoked strong emotional reactions in other viewers — not because it was accurate or useful.

Once you understand this structural incentive, you can adjust your consumption accordingly. Treat social media property content as entertainment, not education. Treat primary data sources — ABS releases, CoreLogic indices, RBA minutes, APRA determinations — as education. The effort required to read a primary source is higher, but the signal-to-noise ratio is incomparably better.

There's one more dimension to this that I think deserves attention: the difference between information and intelligence.

Information is raw data. Melbourne house prices rose 6% last year. Interest rates are at 4.35%. Vacancy rates in the southeast are 1.6%.

Intelligence is information processed through a framework that produces actionable conclusions. Melbourne house prices rose 6% last year AND land supply in established suburbs is constrained AND population growth is running at 2.3% annually, THEREFORE prices are likely to continue rising, THEREFORE buying now is rational for a 5-10 year hold.

The gap between information and intelligence is where most investors get stuck. They have access to enormous amounts of information — more than at any point in history — but they lack the analytical framework to convert that information into decisions.

This is, incidentally, exactly where a qualified buyer's agent adds value. Not by having access to secret information (we read the same CoreLogic data as everyone else), but by having the framework to convert publicly available information into client-specific investment decisions.

The three filters I've shared today are the first step in building that framework. They won't tell you which property to buy. But they'll protect you from making decisions based on information that shouldn't influence your decisions at all. And in a market where the average cost of a mistake is six figures, that protection is worth more than any hot tip.

Why this matters for your actual portfolio

Bad information leads to bad decisions. Bad decisions in property cost six figures.

A client came to us after spending eight months consuming property content on social media. By the time she reached out, she was completely paralysed 8.

We sat down, applied these three filters to every piece of "advice" she'd collected, and threw out roughly 90% of it. What remained was a clear, data-backed case for purchasing a 600-square-metre house in Melbourne's southeast at the $650,000-$700,000 price point, with post-modification rental potential of $800-$850 per week.

She bought. The property has appreciated. The rent is hitting her account. The noise is gone.

High-quality decisions come from a high signal-to-noise ratio. Keep the signal. Dump the noise. And if you're not sure which is which, apply the three filters: Has it actually happened? What's the primary source? Is the advisor consistent over time?

I'm Yan Zhu. I focus on data, not drama. If you'd like to talk property strategy with someone who won't change their thesis next quarter, you know where to find us 9.

I'll share one more real example that illustrates how costly it can be to act on noise rather than signal.

A potential client contacted us in early 2019 in a state of genuine panic. She'd read an article predicting that Victoria's new vacancy tax would cause Melbourne property prices to collapse as foreign investors dumped their holdings. The article had been shared 4,000 times and had 200 comments, almost all expressing alarm.

She'd been planning to purchase a $650,000 house in Narre Warren. She'd done three months of research. She had pre-approval. The numbers worked. But after reading this article, she decided to wait "until the crash."

The crash didn't come. Melbourne house prices in the southeast rose approximately 8% over the following twelve months. The property she'd been looking at sold to another buyer for $670,000 — $20,000 more than she would have paid.

When she came back to us a year later, the entry price for comparable properties had risen to $710,000-$720,000. She'd lost approximately $60,000 in purchasing power because one viral article — which turned out to be completely wrong — overrode three months of careful research.

Noise costs money. Real money. The three filters I've outlined aren't academic exercises. They're financial protection.