MVRV Z-Score and NUPL are the two most widely cited Bitcoin valuation metrics, and they are built from almost the same raw ingredients: market cap and realized cap, the aggregate value the market has assigned to Bitcoin versus what holders actually paid for their coins on average. Given how similar their inputs are, it is worth being precise about how they differ and why a reader might want both rather than either one alone.
How MVRV Z-Score is built
MVRV Z-Score takes the raw difference between market cap and realized cap, unrealized profit across the entire supply in dollar terms, and normalizes it by the standard deviation of market cap over the asset’s history. That normalization is the whole point: it accounts for the fact that Bitcoin’s market cap has grown by orders of magnitude over its life, so a simple ratio would make older cycles look artificially small next to recent ones. The Z-Score puts every cycle on comparable footing, which is exactly why it has been useful for spotting historical extremes. The full mechanics are covered in Bitcoin MVRV Z-Score Explained.
How NUPL is built
Net Unrealized Profit/Loss takes the same unrealized profit figure, market cap minus realized cap, but expresses it as a simple ratio to market cap rather than standardizing it statistically. The result is a metric that moves between roughly -0.5 and 0.75 historically and is typically read through named zones: capitulation, hope/fear, optimism/anxiety, belief/denial, and euphoria/greed. Where MVRV Z-Score answers “how statistically extreme is this relative to Bitcoin’s entire history,” NUPL answers “what fraction of the market’s total position is currently sitting in profit versus loss, and how does that compare to previous emotional phases of past cycles.”
Where they agree
At true cycle extremes, both metrics tend to move together and both have historically flagged the same major tops and bottoms, because they are ultimately drawing on the same underlying imbalance between what the market paid and what it is worth today. A reading that looks stretched on MVRV Z-Score will almost always correspond to a euphoria reading on NUPL, and a capitulation-zone NUPL reading will usually line up with an MVRV Z-Score near or below zero.
Where they diverge, and why that’s useful
Reading them together: a worked example
Imagine bitcoin has rallied hard for six weeks. MVRV Z-Score has climbed but sits mid range, while NUPL has pushed into the belief and denial band. Read alone, one says the market is not stretched by historical standards and the other says holders are sitting on large unrealized gains. Read together, the picture is coherent: unrealized profit is high in absolute terms, but relative to the volatility of this market it has been higher, which is exactly the state a mid cycle advance produces.
Now flip it. Suppose price is flat but the realized price keeps rising, because coins keep changing hands near the current level. MVRV compresses toward one and NUPL drifts toward zero. Neither is signalling danger. Both are describing a market rebuilding its cost basis at a new level, which is usually a healthier setup than a vertical move that leaves the average buyer far underwater or far ahead.
The mistakes that cost people money
The first is treating either metric as a trigger. Both have spent long stretches in what looked like extreme territory while price kept going, and both work far better as context than as timing. The second is comparing today’s readings to cycles from a decade ago without adjusting for how much the market’s structure has changed. The third is using one metric while ignoring supply. Valuation without exchange reserves tells you what the market has paid but not what is available to sell, and that second half is often the part that matters.
The disagreements are informative precisely because of how each metric is built. NUPL’s zone-based framing is more sensitive to shifts in market-wide sentiment over shorter windows, since it directly tracks the profit/loss ratio without smoothing it against long-run volatility. MVRV Z-Score’s normalization makes it slower to flag an extreme unless the move is genuinely large relative to Bitcoin’s full history, which makes it better suited to answering “is this actually a historic extreme” rather than “has sentiment shifted this month.” In practice, that means NUPL will often move into a more excited or fearful zone before MVRV Z-Score confirms the same shift is statistically significant. Reading them together, rather than picking a favorite, gives a faster read on sentiment shifts alongside a slower, more rigorous check on whether the current level is actually unusual by Bitcoin’s own historical standard. Both are best used alongside supply-side context like Bitcoin Exchange Reserves Explained, since a valuation extreme paired with continued supply leaving exchanges tells a different story than the same valuation extreme paired with reserves flowing back in.
Common questions
If they overlap so much, why track both? Because they normalise differently. NUPL is a share of market value, MVRV Z-Score is a standard deviation measure. The divergences between them are where the information is.
Which one is better for spotting bottoms? Historically MVRV Z-Score’s lower band has been the tighter bottom signal, while NUPL’s capitulation zone tends to confirm it. The wider set of signals is covered in our cycle top and bottom indicator guide.
Do these work for altcoins? The maths generalises to any chain with a full transaction history, but the historical bands are calibrated on bitcoin and do not transfer cleanly to assets with shorter or thinner histories.
Leave a Reply