Skip to content

Markets · Bank loans and deposits

Bank loans and deposits, Q2 2026 positions: $22.3tn of cross-border holdings identified by pair; the largest is UK → United States ($2,043bn)

Cite

Cite this view

datasteg (2026), “Market matrix: bank loans and deposits”, datasteg money-flow, release 20261102.2, method 1.5, 2 November 2026, https://money-flow.datasteg.com/en/m/bank?r=20261102.2&d=2026-11-02.

Figure note: Source: datasteg money-flow (release 20261102.2, method 1.5), from BIS LBS.

BibTeX
@misc{datasteg:mf:mkt-bank:20261102.2,
  author       = {{datasteg}},
  title        = {Market matrix: bank loans and deposits},
  howpublished = {datasteg money-flow, release 20261102.2, method 1.5},
  year         = {2026},
  date         = {2026-11-02},
  url          = {https://money-flow.datasteg.com/en/m/bank?r=20261102.2&d=2026-11-02},
  urldate      = {2026-11-02},
  note         = {Positions from BIS LBS}
}
The datasets' own citations (1)
  • Bank for International Settlements (2026), Locational banking statistics, retrieved 30 October 2026, https://data.bis.org/topics/LBS.
The method's citation

datasteg (2026), “Money-flow methodology”, version 1.5, datasteg money-flow, 2 November 2026, https://money-flow.datasteg.com/en/method?v=1.5.

@misc{datasteg:mf:method:1.5,
  author       = {{datasteg}},
  title        = {Money-flow methodology},
  howpublished = {datasteg money-flow, version 1.5},
  year         = {2026},
  date         = {2026-11-02},
  url          = {https://money-flow.datasteg.com/en/method?v=1.5},
  urldate      = {2026-11-02}
}

For each instrument, who holds whose assets · data: Q2 2026 positions · release 20261102.2 · method 1.5 · Updated 2 Nov 2026, 22:57 KST

Filters
Market
Latest 2026-11-02→ newest data by this day: Q2 2026 positions
How to read this
  • Direction. Rows hold, columns issued. One cell reads: “Banks located in row hold the value of loan and deposit claims on column residents”.
  • A stock, not a flow. Values are Q2 2026 positions, not what was invested that year; they grow with prices even without transactions.
  • Residence basis. A holder is where the position is booked. A holder in an FC or FD economy is often not the final investor.
  • FC · FD. FC is an offshore financial centre: An offshore centre in the BIS grouping: shell companies, special purpose vehicles, hedge funds and holding companies of foreign firms (e.g. Chinese firms incorporated in the Cayman Islands). A holder here is usually a pass-through, not the final investor. FD is a fund domicile: Where investment funds are legally domiciled (Luxembourg, Ireland, the Netherlands, Switzerland). Holdings booked here are mostly the assets of funds that investors elsewhere bought.
  • Gross totals double count. The same money can appear twice: UK investors hold $204bn in Luxembourg (UK row × Luxembourg column), and Luxembourg holds $172bn in Germany (Luxembourg row × Germany column). Money can also return home through a pass-through, as in France ⇄ Ireland (round-tripping).
  • Big markets are big on both sides. This is gross data, so large markets are large as holders and as destinations; switch to Net to see the difference.

How this differs from the Ladder: the panel tab Ladder vs Markets. For holdings restated by nationality (the final investor): Coppola, Maggiori, Neiman and Schreger (2021).

Bank loans and deposits, Q2 2026 positions · cells: $bn

Cell value (log bins)$10bn$30bn$100bn$300bn$1tn$3tndarker = largerbrighter = larger
■ square area ∝ valueNumbers: cells of $30bn or moreDomestic, not cross-border (left out)Totals and Other in grey, off the scaleFCOffshore financial centreFDFund domicile
Holder ↓ · issuer →
Total · share outside the table (Other)
2,043
213
453
350
356
297
156
204
109
120
$5.7tn · Other 23.8%
1,240
1,103
160
151
340
452
127
65
77
31
$4.6tn · Other 17.6%
483
441
179
290
255
57
176
153
52
329
102
$3.3tn · Other 23.1%
92
115
58
112
119
$1.2tn · Other 51.1%
761
151
72
61
$1.1tn · Other 2.9%
126
161
152
81
61
58
56
$1.0tn · Other 22.8%
114
112
115
70
68
97
$0.8tn · Other 21.2%
36
159
172
34
$0.7tn · Other 27.1%
41
87
158
72
34
$0.5tn · Other 15.2%
85
80
34
$0.5tn · Other 33.1%
61
82
33
47
$0.4tn · Other 23.5%
38
40
93
57
35
$0.4tn · Other 26.1%
Column total (hatched = holders outside the table)
7.0%
12.6%
2.5%
4.9%
7.4%
6.0%
1.8%
5.0%
8.2%
17.5%
2.0%
4.4%
Identified total $22.3tn

Source: BIS LBS (Q2 2026). Calculations: datasteg money-flow, method 1.5, rules R1, R2, R5, R11.

Cut a lender: who loses, against net reserves

A what-if on this matrix (rule R21). Pick a lender and every debtor it lends to takes the cut, or pick a debtor and each of its lenders cuts. Loss = cut × the lender’s share × the debtor’s bank claims from all BIS reporters, shown against the debtor’s net reserves (the Signal Board’s A2 denominator). A scaling of reported positions, not a forecast: banks also roll over, hedge and replace.

Cut by
30%

If each lender cut its claims on Korea by 30%, the largest loss would come from banks in Hong Kong: $11bn, 2.6% of Korea’s net reserves ($434bn).

  1. Hong Kong$11bn23.3%
  2. UK$5bn10.3%
  3. United States$5bn10.0%
  4. France$2bn4.7%
  5. Taiwan$1bn2.8%
  6. Macao$1bn2.6%
  7. Australia$0bn0.8%
  8. Luxembourg$0bn0.5%
  9. Spain$0bn0.4%
  10. Netherlands$0bn0.3%
  11. Belgium$0bn0.1%
  12. Canada$0bn0.1%
  13. Italy$0bn0.1%
  14. Philippines$0bn0.1%
  15. Sweden$0bn0.1%

Source: BIS LBS; net reserves from the Signal Board (IMF IRFCL) · Positions Q2 2026 · Rule R21

datasteg: the flow sites