One theme, five tokens, two thirds of the volume in one of them: attention inside the memory trade

TL;DR

This is the most heavily funded theme on the book, and almost all of the attention inside it sits on one name.

• SNDKB did $97.80M in 24 hours, the largest of all 100 instruments on the book, while ranking only 6th by market cap ($76.98M). It is 11.20% of the whole book volume on its own.
• The five tokens did $147.51M together and SNDKB is 66.3% of that. The other four add up to $49.71M, so SNDKB alone is 97% more than their combined total.
• Remove SNDKB and the group stops being unusual. Group turnover is 0.7723x; without SNDKB it is 0.4360x, below the 0.6625x of the whole book. The gap between this group holding 14.5% of the value and doing 16.9% of the volume comes entirely from one instrument.
• The thematic ETF is the coldest large instrument on the book. DRAMB holds $19.64M, 14th by market cap, and turns over 0.186x, 96th. SNDKB turns 6.83 times faster, does 26.9 times the volume, and is only 3.9 times larger.
• But "single names trade more than baskets" fails once market cap is held constant: AVGOB is a single stock worth $17.80M — almost exactly DRAMB size — and turns over 0.224x against DRAMB 0.186x. The real regularity is that large positions turn slowly, not that baskets do.
• Four of the five fell further than the book median over 30 days (−35.29% / −20.60% / −21.49% / −20.43%); only MUB at −16.31% did slightly better than the −16.71% median. This theme fell harder than the book this month.

1. What these five are

bStocks is Binance's in-house tokenized-equity issuer (BTech Holdings Limited, ADGM / FSRA listed certificates, BNB Chain only, traded on Binance only). CoinMarketCap reports 105 instruments; the snapshot behind this piece holds 100 of them, timestamped 2026-10-31T06:06:52Z.

These five are the whole memory and storage line: three NAND / DRAM makers (SanDisk, Micron, SK Hynix), one drive and enterprise storage maker (Western Digital), and one ETF that covers exactly this theme (Roundhill Memory ETF). The same supply chain is offered both as single names and as a basket, which is rare in this snapshot.

The group holds $190.99M, 14.49% of the book, and did $147.51M, 16.89% of its volume. That is the largest market-cap share of any group here.

Two notes before anything else. SKHYB tracks a company listed in Korea, not in the United States — a sector called tokenized US equities carrying Korean names is a classification problem, not a data error, and the previous piece raised the same point about SKUUB. Also, SNDKB, MUB and SKHYB all appeared in that piece, there playing the role of spot counterparts to three 2x leveraged tokens, where the finding was that the spot side did more absolute volume. This piece does not repeat that comparison — the contrast here is inside the theme, not against a leveraged wrapper.

2. The largest volume on the book is in this group

The five largest 24h volumes on the book:

SNDKB is first, and it is only 6th by market cap. Two of the five are larger than it (SPCXB, CRCLB), both flagship instruments from the first piece in this series, turning over 0.458x and 0.326x — their volume is a product of their size. SNDKB volume is not.

Inside the group:

SNDKB is 66.3% of the group volume on 40.3% of its market cap. The other four did $49.71M between them, so SNDKB alone is 97% above that combined total.

Strip it out and the group changes character:

The remaining four turn over 0.4360x, under two thirds of the book figure. So the accurate version of "the memory line is being traded hard" is "the SanDisk name is being traded hard".

3. The thematic ETF is the coldest large instrument on the book

DRAMB is something this snapshot rarely offers: an ETF covering the exact industry the other members sit in, wrapped as a token by the same issuer. Its $19.64M makes it the 14th largest instrument on the book — slightly under TSLAB ($23.68M), slightly over SPYB ($19.08M), squarely in the upper middle of this snapshot.

Its turnover of 0.186x ranks 96th. The five lowest turnovers on the book are DRAMB (0.186x), CRMB (0.183x), GMEB (0.160x), COHRB (0.141x) and AMATB (0.130x) — the other four hold $501.9K, $6.61M, $8.44M and $9.02M, so DRAMB is 2.2 times the largest of them.

Next to SNDKB:

Volume differs 26.9-fold, market cap only 3.9-fold, so turnover differs 6.83-fold. This runs in the opposite direction to the denominator trap in the previous piece: there the numerators were close and the denominators differed 27-fold; here the denominators are the same order and the numerators differ 27-fold. One ratio, two completely different sources — which is why turnover has to be read next to absolute volume.

DRAMB is also the coldest of the six ETF tokens on the book:

DRAMB turns over 43% slower than the second lowest, QQQB, whose market cap is 5.0 times larger. On a "large positions turn slowly" logic, DRAMB should have been the faster of the two.

4. The counter-example: hold market cap constant and the single-name story collapses

That 6.83-fold gap reads easily as "this channel buys names, not themes". Two findings in this snapshot contradict it directly.

Counter-example one: AVGOB is a single stock, almost exactly DRAMB size, and turns over almost exactly as slowly.

Broadcom is the second largest chip designer in the world and an unambiguous single name. Market caps differ by 10%, turnover by 20%, and both sit at the bottom of the book. If "nobody wants baskets" explains DRAMB at 0.186x, it does not explain AVGOB at 0.224x.

Going further down: among the ten lowest turnovers on the book, five hold more than $5M — AMATB ($9.02M / 0.130x), COHRB ($8.44M / 0.141x), GMEB ($6.61M / 0.160x), DRAMB ($19.64M / 0.186x) and AVGOB ($17.80M / 0.224x). Four of the five are common stocks. The actual regularity is stated in the issuer note itself: the five largest instruments on the book turn over 0.38x against 0.92x for the other 95 — large positions turn slowly, regardless of wrapper.

Counter-example two: WDCB does less absolute volume than DRAMB.

These two listed on the same day (2026-07-06), an identical 117-day window:

WDCB turns over 4.75 times faster than DRAMB and still does 36% less absolute volume. Western Digital is one of the four single names on this line, a storage company with revenue in the tens of billions — and its token here holds $2.63M, 48th on the book. For WDCB, "single names trade more than baskets" holds as a ratio and fails as an amount.

So the claim has to be narrowed to something that survives: not "names beat baskets", but "the attention in this theme sits on one name". Outside SNDKB, MUB turns over 0.415x, below the book average; DRAMB at 0.186x is near the book minimum; and WDCB cannot even match DRAMB on volume. Only SanDisk is traded repeatedly.

5. Range width: a 1.6-fold spread inside one theme

Range width is the listed high divided by the listed low. Among the 46 instruments listed 100 days or more, these five sit here (1 = narrowest):

The widest, SNDKB, is 1.58 times the narrowest, DRAMB. That ordering is what the structure predicts: the thematic ETF spreads the idiosyncratic risk of five companies, so it is narrower than any single name in the set; and 1.543x is still above the 1.350x median width of ETFs on the book (n=4), which says the theme itself kept moving after the spreading.

One reading worth holding on to: SNDKB sits at 13.0% of its own range, the lowest of the five, −51.2% from its listed high. First on the book by volume while cut in half from its own peak — where the volume concentrates and where the price sits are two independent facts.

Among those 46 instruments, the median width for common stocks is 1.768x (n=28). The three members with long enough windows print 2.432x / 2.005x / 1.767x, so two are above the median and MUB sits exactly on it. On width this theme is not an outlier at all; its distinguishing feature is only the concentration of volume.

6. Putting this group back in the book

First, listing early matters more than anything else. The book by listing month:

The ten earliest instruments hold 52.3% of all book value; the fifteen listed in the last month hold 0.9%. SNDKB and MUB are two of those ten, worth $139.60M together, 20.3% of the June cohort. The other three members belong to the July cohort. Half the reason this group is large is that it listed early.

Second, this theme fell harder than the book this month. The 30-day median across the 100 instruments is −16.71%, with 82 negative. All five members are negative and four fell further than the median: SNDKB −35.29%, DRAMB −21.49%, WDCB −20.60%, SKHYB −20.43%; only MUB at −16.31% did slightly better.

Third, the 7-day window points the other way. SKHYB +9.55%, WDCB +7.44%, DRAMB +2.68% and MUB +2.18% are all positive, and only SNDKB at −6.84% is negative — the one instrument the volume concentrated in is the only one of the five down over a week. Its 24-hour move, −10.44%, is also the largest in the group.

Fourth, the kind mix. The book holds 67 common stocks, 8 unclassified, 6 ADRs, 6 ETFs, 10 leveraged ETFs, 2 inverse ETFs and 1 pre-IPO. This group is four common stocks and one ETF, the most ordinary composition available in this snapshot — no leverage, no ADRs, nothing unclassified. Every distinguishing feature of this group comes from the volume distribution, none from product structure.

7. Structure: what these five tokens legally are

All of this comes from the issuer record on this site, the same data behind /stocks/bstocks:

• Legal wrapper: tracker certificate, ADGM / FSRA listed, issued by BTech Holdings Limited. The holder gets a certificate that tracks a price, not a share registration.
• Voting: none.
• Dividends: pass-through, but a 30% US withholding deduction comes first, and the remainder is reinvested through a multiplier.
• Redemption: available, KYC required, $5 minimum.
• Proof of reserve: attestation, not a live on-chain proof.
• US persons: blocked.
• Chain: BNB Chain only. BSC deposits only, so a wrong-network transfer loses the asset.
• Issuance model: exchange self-issued; the only venue is Binance, operated by the issuer.
• Grade on this site: B.

Two of those matter extra for this group.

DRAMB is a wrapper on top of a wrapper. Roundhill Memory ETF is itself a fund, and the fund holds the stocks; the certificate tracks the fund price; the token represents the certificate. Three nested layers, with the costs of the middle two borne by the token holder — the ETF management fee comes out of fund NAV, and the certificate layer cost structure is not in this data. The four single-name tokens have two layers.

"Dividends pass through" means something different for each of these five. Some of the makers pay a dividend, so the "30% withholding then reinvest through a multiplier" rule genuinely applies to them; DRAMB tracks an ETF, which receives distributions from its constituents first and then decides what to pass to its own shareholders, so the token layer receives something that has already been processed twice. The current multiplier is not in this data — see section 9.

8. This snapshot was taken on a Saturday

The timestamp 2026-10-31T06:06:52Z is a Saturday morning UTC. US cash equities closed on Friday at 20:00 UTC and do not reopen until Monday — for more than two days after this snapshot, these token prices are the only prices these names have.

There is an extra layer for this group: two of the five underlyings do not trade in the United States at all. SK Hynix is listed in Korea, and Seoul hours do not overlap US hours; the ETF behind DRAMB trades in the United States, but its constituents span US, Korean and Japanese markets. "Closed" is not the same moment for all five of these tokens.

This has already been quantified once on the same issuer: one stock, four books lays the four tokenized versions of AAPL side by side on a Saturday and concludes that a quote outside trading hours is a market maker opinion, not a market price. The 24-hour window in this piece runs from Friday 06:06 UTC to Saturday 06:06 UTC and contains one complete US session, so it is not a pure closed-market quote; but part of that $97.80M in section 2 fell in hours when US equities were already shut.

9. What this method cannot compute

First, no premium or discount. The snapshot holds only the token price, with no reference price for the underlying stock or ETF. More to the point, the bStocks dividend mechanism is "deduct 30% US withholding, then reinvest through a multiplier" — without knowing the current multiplier, the token unit price and the underlying share price are not comparable in the first place. So no premium figure appears here.

Second, no shares per token. The snapshot carries no conversion ratio. SNDKB has 66,641 tokens outstanding at $1155.20 each, and neither number converts into a share count without that ratio — a token price far above any of these companies actual share prices says the ratio is not 1:1, but this data does not say what it is.

Third, nothing about the fundamentals of this theme. This snapshot holds no company financials and no memory spot prices. When this piece says "the memory line", it means the trading behaviour of these five tokens in this channel, not a position in the industry cycle. Any judgement about where a supercycle stands is outside what this data can support.

Fourth, the range-width windows are unequal. These five are 142/142/114/117/117 days. The book comparison uses only the 46 instruments listed 100 days or more, and even there the windows differ by more than a month — the longer something is listed, the more easily it reaches an extreme.

Fifth, market cap is CMC's definition (price times circulating supply), and the book total of $1,318.38M is the sum of these 100 rows; the bStocks row on this site's /stocks page shows an on-chain size of $914M — two figures from a different source and a different moment, which will not tie out. This piece uses the former throughout.

Sixth, 100 of 105. CMC reports 105 instruments; the snapshot holds 100. The missing five appear in no figure here.

Seventh, turnover is 24h volume divided by market cap. A small denominator inflates the ratio mechanically, so high turnover in the tail is first of all a consequence of a small market cap and must not be read as better liquidity in the tail — the two counter-examples in section 4 are a direct consequence of that rule.

Conclusion. Memory and storage is the most heavily funded theme on bStocks: five tokens holding 14.49% of book value and doing 16.89% of its volume. But the gap between those two shares comes entirely from one instrument — SNDKB did $97.80M, the largest 24h volume of all 100, 66.3% of this group, and 97% more than the other four combined. Take it out and the remaining four turn over 0.4360x, a third below the book 0.6625x.

And "the single name trades 6.83 times faster than the thematic ETF" does not survive a control: AVGOB is a single stock at almost identical market cap and turns 0.224x against 0.186x, both at the bottom of the book; WDCB is a single stock and does 36% less absolute volume than DRAMB. The main explanatory variable for turnover is position size, not wrapper — and the issuer own data says the same thing: the five largest instruments turn 0.38x, the other 95 turn 0.92x.

The sentence to carry out: this channel did not put money into a memory theme, it put money into the SanDisk name. For the other four tokens, trading activity bears no relationship to their standing in the industry.

Source CoinMarketCap, timestamped 2026-10-31T06:06:52Z, covering 100 of the 105 instruments bStocks reports. Instrument detail and quotes at /stocks/bstocks. Research notes, not investment advice. DYOR.