So. In a world just flooded with information, right? It feels like we're drowning in data sometimes. How do you actually cut through all that noise to get, you know, truly valuable forward-looking insights about the economy?

I mean, how do you really know what's coming before everyone else does? It often feels like trying to drive using only the rearview mirror, doesn't it? A foggy one at that.

Well, today in the deep dive, we're going to explore this really fascinating collaboration. It's like they're upgrading that foggy mirror to a crystal clear predictive radar for the economy. We're talking about the partnership between LC, that's the London Stock Exchange Group, and a company called Extinential Technologies or Extech. They've teamed up. And they're delivering global macro forecasts in a way that seems to be earlier and often far more accurate than what we traditionally rely on. It feels like a genuine shift in how economic foresight is actually.

Achieved. It really is. And these macroeconomic forecasts, they're just absolutely crucial, not just for the big financial institutions making, you know, multibillion dollar investment choices, but for you, too. For anyone just trying to understand the economic winds, how they might affect your daily life, your finances, traditional methods, well, They have their place, but they often lean on lagging indicators or surveys, and that inherently limits how early and how precisely you can see what's around the corner. This deep dive, I think, will show how this new partnership is tackling those exact limitations, offering something genuinely different.

Okay, so let's unpack this. Who are these players, the architects behind this economic foresight engine? And what makes their alliance so potent? Let's start with exponential technologies, ex-tech. The CEO is Morgan Slade and his background, well, it's pretty remarkable. We're talking 25 years of institutional investing experience, a real pioneer in using AI tools to pull out vital investment signals from huge data sets. He's researched, he's traded all the major global asset classes.

So he's a great guy. Deep practical market knowledge, not just theory.

Plus, he's got MIT engineering degrees with a finance specialty. That really underscores the technical side.

Yeah. And.

What's truly fascinating, I think, is how Extech blends that really deep understanding of institutional investing, gained over decades right there in the market trenches with absolutely cutting edge data technology and advanced analytics. They've sort of specialized in turning complex raw data into signals you can actually use. They build innovative data products, high performance delivery systems, all aimed at institutional investors. It's really about bringing that scientific rigor to market.

Prediction. Okay.

And then the other half of this powerhouse duo. Elseg, the London Stock Exchange Group. Their contribution sounds absolutely fundamental. You mentioned world leading historical point in time data, PT data and robust consensus economic estimates. Now, this PIT data, maybe unpack that a bit. Why is it such a big deal?

Right. The PIT data. Think about it like this. Traditional economic data you often see. It's like looking at a history book that's been edited and updated over time. Numbers get revised. But El Seg's PIT data. That's like having the original newspaper from that day. It shows you the exact data as it first appeared at any point in time you choose. It includes the initial releases and all the later revisions.

Okay. So when you're building models.

Exactly. When you're building predictive models, you can completely eliminate what's called revision bias. You're using the actual information the market had at that specific moment. You get a true unvarnished picture of what people were reacting to in real time. It's a massive advantage for modeling accuracy.

So it's like looking over the shoulder of the past, seeing what they saw then.

Precisely. And if you connect this to the partnership, the synergy is clear, isn't it? It's like a really well-designed machine. X-Tech brings the super intelligent brain, the advanced analytics fuel.

Engine. And LSEG provides the.

The incredibly rich, historically accurate fuel and the data map. So together, they're not just guessing where things are going. They're modeling the journey with this unprecedented precision. And their whole mission really is to build these predictions from the bottom up based on completely independent sources.

Which makes it different from the usual forecasts.

Totally different. They call it orthogonal, meaning it's coming from a completely different angle, a different fundamental perspective than, say, traditional broker forecasts. It's not just an incremental improvement. It's more like a reimagining of the process.

Right, orthogonal. If they're building this sophisticated economic radar, what exactly are they predicting with it? And how do they manage to get such a, well, such a jump on everyone else? What did they tackle in their first release?

Yeah. Well, they're covering some of the really big U.S. Macroeconomic announcements, the ones that really move markets. We're talking the U.S. Consumer Price Index, CPI, you know, measuring changes in prices we all pay. Then there's the Michigan Consumer Sentiment Index and the Conference Board Consumer Confidence Index, both. Gauge consumer mood via surveys. And.. US retail sales, measuring total spending at retail businesses. These are the big ones.

And what's really insightful is they don't just give you the headline number, the top line forecast for these. They go deeper. Much deeper. They offer granular forecasts for individual CPI categories.

So imagine breaking CPI down into things like gasoline, food, shelter, transportation services, and even more specific things like medical costs, education, recreation, spending.

Wow. Okay, that level of detail must be critical.

It is. Take gasoline, for example. It's often Stripe out of core inflation because it's so volatile, right? Right.

So being able to forecast that accurately as its own component is incredibly valuable. If you're trying to really understand the nuances of inflation, it just allows for much more informed strategic thinking.

Okay. The real genius then must be in how they pull this off, this bottom-up prediction, especially for something as complex as CPI. How does that process actually work? And importantly, how do they make sure it's not just some AI black box? Is it explainable? That's.

A great point. And no, it's definitely not a black box. They're very transparent about the process. They follow this rigorous five step method for developing the forecast. CPI is a good example to walk through.

So first, step one, they identify surrogate CPI features.

Surrogate features, meaning?

Meaning they pinpoint specific economic indicators that actually influence the individual CPI components. Things you might expect, like average hourly earnings, private payroll numbers, unemployment, But proprietary data they found useful, like used car prices, surprisingly. They found that correlates really strongly with certain parts of CPI. It's about finding the real drivers underneath.

Interesting. Okay, so find the drivers. What's.

Next? Step two is gather and prepare data inputs. This means collecting all that relevant data, LSG's PIT historical data, their own proprietary alternative data, using AI to process it all. And they do something called feature engineering.

Feature engineering. Sounds technical. It.

Is a bit, but the idea is that It's simple. It's about transforming the raw data into new insights, new features that are more powerful for prediction, making the data work harder, basically finding hidden patterns. Then, step three is to forecast individual CPI components.

So they model and predict each piece separately, gasoline, food, shelter, using tailored approaches for each Exactly.

One. Okay, so tailored models for each component.

Then, fourth step. Evaluate and refine models. This isn't a one and done thing, it's ongoing. They constantly analyze how each model is performing, tweak it, iterate, make improvements, keeps the forecasts robust.

Makes sense. Continuous improvement.

Right. And finally, step five. Aggregate individual forecasts. They combine all these carefully built imponent forecasts, the gasoline forecast, the food forecast, etc. To produce the overall CPI month over month percentage change forecast.

So you see, it's built block by block, not a black box, but more like a meticulously engineered prediction engine building the big picture from the ground.

Up. That does sound incredibly rigorous, building it piece by piece. But with all those inputs and steps. What's maybe the biggest challenge, the potential pitfall they have to constantly watch out for in that process?

Yeah, good question. I'd say the biggest potential pitfall is always data quality and just the sheer dynamic nature of economic relationships. Things change. Correlations.

Shift. So how do they guard against that? And what's the secret sauce you mentioned behind that constant improvement, the refining of models? How do they keep it so current?

They tackle that dynamic nature head on. Continuous validation is key, of course, but a core part of their methodology is a modeling technique called teacher forcing. Teacher forcing.

Yeah. Or sometimes called one step ahead forecasting. Imagine teaching a student to predict a sequence of numbers. Okay. Instead of letting them make a guess for the next number, then wait and see if they were right way down the line. Teacher forcing means you immediately show the model the actual outcome for the step it just tried to predict.

So it gets instant feedback.

Instant feedback, exactly. It learns and refines its approach based on the latest real-world data right away. It makes the model constantly adaptive, highly responsive to new information. This relies heavily on machine learning, advanced statistical modeling, processing huge amounts of public and proprietary data. It allows for those dynamic adjustments and helps maintain that accuracy.

Okay, teacher forcing, constantly learning from reality. So the methodology sounds robust, very sophisticated. But what about the bottom line, the results? How do these forecasts actually perform out in the wild, especially when you compare them to, you know, the standard consensus estimates we usually see? That's the real test, isn't It?

Absolutely. And this is where it gets really compelling. The key differentiator here is that these forecasts are consistently both earlier and more accurate than those traditional economist consensus estimates. Let me give you some numbers. These are averages from backtesting November 2017 through April 2025. Okay. For headline CPI. Their first and second forecasts. They show a stunning 99.99% correlation and 92% directional accuracy for the index level itself.

Well, 99.99% correlation, that's almost perfect alignment.

It's incredibly high. And even when you drill down into CPI gasoline, which as we said, is notoriously volatile, it still shows 99.8% correlation and 92% directional accuracy for that specific category forecast.

Still incredibly strong for something so tricky. Very.

Strong. And it's not just inflation. For something like the Conference Board Consumer Confidence, they achieve 88% correlation and 70% directional accuracy.

So reliable predictive power across different kinds of.

Indicators. Okay, those accuracy numbers are impressive. But you stressed the earlier part. This is where it becomes a potential game changer, right? How much earlier are we talking? Days? Weeks? And does this accuracy mean they almost never miss? Or are there certain economic conditions where even this system, finds it harder to predict.

We are talking significant lead times. And look, no system is absolutely perfect 100% of the time, especially in economics. Unexpected shocks can happen. But the consistency is remarkable. Let's take CPI again, the first CPI forecast. It comes out on the third Monday of the current number.

Month. The current month.

So way before the official.

Way before. Nearly a full month specifically. 25 days ahead of the official Bureau of Labor Statistics release. And crucially, this is often before many traditional brokers even put out their initial predictions.

25 days ahead. What's the accuracy like that far.

Out? Even that early, it shows an 82% correlation with the final number, 75% directional accuracy, meaning it gets the direction of the change right there. Three quarters of the time and 94% sign accuracy, meaning it gets, whether it's positive or negative, right almost all the time.

That's still pretty impressive for being almost a month early.

It is. Then they release a second CPI forecast. This one comes out around a week before the official release date specifically, on the third trading day of the following month. This one incorporates additional, more recent data, so the accuracy ticks up even further. You're looking at 84% correlation and 80% directional Exactly.

Accuracy. Okay, so an early look and then an even sharper look closer to the date.

And overall, if you average it out across their forecasts, these LGX tech predictions are available an average of 12 days before the professional economist's consensus estimates typically solidify. Which is why their CPI forecast is considered, and I quote, "the most accurate CPI forecast in the world." world 25 days before the release. That's a pretty bold claim, but the data seems to back it up.

12 days on average. 25 days for that first CPI look. That is a significant timing advantage.

So what does this all really mean for how we usually get our economic information? It sounds like a fundamental shift. It almost makes you wonder how much past market movement was just, you know, noise reacting to imperfect late information, doesn't it? It.

Absolutely does raise that question because unlike traditional forecasts, which, as we said, often lean on surveys, lagging indicators, maybe models that aren't updated as frequently and are largely broker driven estimates, LSEC and Extech are leveraging these high frequency, often real-time, time data streams, and constantly updating their models with techniques like teacher forcing. This provides what they rightly call a transformative informational advantage. It really is about getting that edge of foresight, not just hindsight, moving from reacting to anticipating.

Foresight, not hindsight. I like.

That. For.

Investors, yeah, it changes the game. So let's connect those dots directly for the listener. Beyond the impressive stats in the cool tech, why should someone listening right now care about these advanced macro forecasts? What's the real world relevance? The application of this knowledge, especially if you're involved in any kind of investment or economic strategy, whether big or small?

This edge of foresight, as we called it, allows institutional investors, the big players, to make much more informed and crucially timely decisions. Imagine being able to, say, position your portfolio strategically weeks ahead of a major macroeconomic release instead of scrambling to react after the number hits the Exactly.

Wires. Right, being proactive instead of reactive.

It allows them to enhance their trading strategies. They can integrate these forecasts into directional bets or relative value trades, betting one thing will outperform another based on the forecast. Looking for those unique opportunities. Furthermore, they can potentially arbitrage market mispricing. If the consensus estimates are way off from these more accurate earlier forecasts, there's an opportunity there before the rest of the market catches up.

Capitalizing on the gap between perception and reality. Precisely.

And of course, managing risk exposure. If you anticipate a shift in inflation or growth or consumer spending, you can adjust your portfolio allocations proactively before the market moves against you. It helps optimize asset allocation across different classes, equities, fixed income, commodities, foreign exchange, aligning your bets with where the predictive trends are pointing. And it just improves market timing for entering or exiting positions.

Okay, so better timing. Better risk.

Management. And maybe most interestingly, it allows them to anticipate surprises and disappointments relative to what the general market expects. If you have high confidence that the consensus is wrong, you can pre-position yourself.

Sometimes this means being a liquidity provider during those brief market shocks when everyone else is panicking.

Stepping in when others step back.

Right. Providing capital when it's needed, which can stabilize markets and be profitable rather than just being caught flat footed by the surprise. And even if you listening aren't managing billions, understanding these early signals can still indirectly empower your own financial planning. How so?

Well, think about anticipating changes in your retirement portfolios value before they happen or understanding potential shifts in the job market based on economic direction or even timing major purchases better based on where consumer confidence is. Confidence might be heading. The ripples spread out.

That makes sense. It's not just about knowing what's happening right now. It's about having a much clearer idea of what's likely to happen next and understanding the implications across the board. "Moving from reacting to genuinely predicting economic shifts." So to wrap up, this partnership between LSAG and Exponential Technologies, it really does seem to be redefining macro forecasting. By combining that deep institutional expertise, the cutting-edge AI, and this comprehensive data, both historical PIT and real-time streams they're delivering earlier, more accurate and, frankly, far more actionable insights into the economy. It feels like a significant leap forward.

It truly does. And as we think about this... This increasing precision, this growing lead time in economic forecasts, it leaves us with a pretty provocative thought, doesn't it? Consider how this might not only change how financial markets operate, which it clearly will, but could it start to influence broader economic policy decisions if policymakers have clearer foresight? And maybe even for us as individuals, how might this growing clarity change how we think about our own financial futures? What new opportunities, but maybe new challenges, might arise when the economic radar becomes just that little bit clearer? Clear for everyone.

Something to mull over.