Eight funds can still be one bet#
Owning many things is not the same as owning different things. A common portfolio holds an S&P 500 fund, a total-market fund, a growth fund, and a technology fund. Four tickers, four expense ratios, one bet: the same handful of mega-cap stocks dominates all four. When those names fall together, the whole portfolio falls together, and the owner discovers that "diversified" was a feeling, not a fact.
Diversification only works when your holdings respond differently to the same events. Measuring that requires looking through the fund wrappers to what is actually inside. That is the job of the Allocation page.
Asset classes: the first cut#
The broadest split is by asset class. Pyx classifies holdings as Equity, ETF, Fixed Income, Cash, or Other. Asset classes are the first cut because they respond to different forces. Stocks price in growth expectations. Bonds price in interest rates and credit. Cash prices in nothing, which is precisely its role: it cannot fall.
The Allocation page shows this split as a donut, a bar chart, or a treemap, with a legend carrying each category's dollar value and weight. When cash distorts the picture (a large settlement balance, say), the Exclude cash toggle drops it and recalculates the remaining slices to 100%, so you see your invested mix on its own.
Sector lookthrough: what your funds actually hold#
A fund is a container, not an exposure. Sector Allocation opens the containers: Pyx decomposes your stock and fund money by industry sector, so a broad-market ETF stops being one slice and becomes its real mix of technology, financials, health care, and the rest. Anything Pyx cannot classify lands in an explicit "Unclassified" bucket rather than being silently dropped, so the chart never pretends to more precision than the data supports.
This is where the eight-funds-one-bet problem becomes visible. If technology reads 40% of your sector chart while you thought you owned "a bit of everything," the lookthrough just told you something your account statements never would.
Concentration: the quiet 20% position#
Weight is a holding's share of your total portfolio value. Concentration risk is what happens when one weight grows large enough that a single company's bad quarter becomes your bad year.
The dangerous part is that concentration happens quietly, without a single decision. A winner that compounds faster than the rest of the portfolio takes over by itself. A position bought at 8% of the portfolio can sit at 20% three years later, and nothing about your daily routine flags it. The Holdings table's Weight column and the Allocation treemap both make this visible at a glance. There is no universal threshold, but once a single stock passes 10 to 20% of the total, you are no longer diversified against that company, whatever else you own.
Target weights and drift#
A target allocation turns "diversified" from a feeling into a number. You decide the mix you want, say 60% equity, 30% fixed income, 10% cash, and enter it under Settings, then Allocation. From then on the Drift column shows the gap between what you hold and what you decided: a drift of +2.4% on equity means equities are 2.4 percentage points over target.
Drift is not a malfunction. It is what markets do to every allocation, every day. Rising assets grow their own weight, so a portfolio drifts toward whatever has recently done best, which quietly increases risk exactly when prices are highest. When total drift crosses your threshold, 2% by default, Pyx highlights the summary so the divergence does not go unnoticed.
Pyx shows the drift. It does not place trades or recommend what to buy or sell.
Correlation: the test of a real hedge#
Two holdings diversify each other only if they do not move together. That is measurable. Correlation runs from +1 (they move in lockstep) to -1 (they mirror each other), and the useful diversification lives near zero or below. Many pairs that look different on paper, a US fund and an international fund, for instance, run correlations high enough that they behave as one holding in a drawdown.
The chart workspace's Rolling Correlation indicator plots this over time for any two symbols, which matters because correlation is not constant: pairs that diverge in calm markets often converge in crashes. See volume and correlation for how the calculation works and how to read it.
Rebalancing: selling winners on purpose#
Rebalancing means trading back toward your targets: trimming what grew, adding to what shrank. It feels backwards, because it always means selling whatever has been performing best. That discomfort is the point. Rebalancing is a mechanical rule that forces you to take risk off after it has paid, and add it where prices have fallen, without requiring a forecast or a feeling.
A practical pattern is to rebalance on drift rather than on the calendar: act only when a category exceeds your threshold, which is exactly what the drift display and banner surface. Two honest caveats. In a taxable account, selling winners realizes gains, so check the cost basis consequences before trading, and consider directing new deposits toward underweight categories instead, which rebalances without selling. And Pyx will not do this for you. It shows the gap and leaves the decision where it belongs.