IPv4 Transfer Analytic View

IPv4 Transfers
Analy'c View
Alain Durand, February 24th 2016

| 2
Questions For This Study
A.  IPv4 Transfer Market Health
1)  What is the concentra'on of address holders?
2)  Is the transfer market dominated by a few buyers?
3)  Is there a regional direction of transfer?
4)  What is the size distribution of transferred blocks?
5)  How are things changing over time?
B.  Impact on the routing table
Do transfers impact the size of the routing table?
C.  Registry Accuracy
Does the WHOIS database accurately reflect who controls
resources?

| 3
Herfindahl-Hirschman Index (HHI) Source: Investopedia.com
Investopedia:

The Herﬁndahl-Hirschman index (HHI) is a commonly accepted measure of
market concentra=on. It is calculated by squaring the market share of each ﬁrm
compe'ng in a market, and then summing the resul'ng number.

The HHI number can range from close to zero to 10,000.

The U.S. Department of Jus'ce considers a market with
a result of less than 1,000 to be a compe==ve marketplace;
a result of 1,000-1,800 to be a moderately concentrated marketplace;
and a result of 1,800 or greater to be a highly concentrated marketplace.

| 4
A.1) Share of the Top 10 Address Holders per
Region: 01/01/2016: ARIN, APNIC & RIPE
Total: 1.7 billion
Top 10: 32%
HHI = 248
Total: 863 million
Top 10: 32%
HHI = 147
Total: 802 million
Top 10: 20%
HHI = 61
ARIN RIPEAPNIC
Source: delegated-extended files from ARIN, APNIC & RIPE
Top 10
Others

| 5
Total: 180 million
Top 10: 44%
HHI = 289
Total: 87 million
Top 10: 44%
HHI = 262
LACNIC AFRINIC
A.1) Share of the Top 10 Address Holders per
Region: 01/01/2016: LACNIC & AFRINIC
Source: delegated-extended files from LACNI & AFRINIC

| 6
A.2) IPv4 Address Transfer Recipients: 2014/2015
Total: 38.9 million
Top 10: ??%
Total: 10.2 million
Top 10: 61%
Total: 18.3 million
Top 10: 42%
RIPEAPNIC
Source: ARIN, APNIC, RIPE: APNIC & RIPE transfer stats files
RIPE does
not report
on
“legacy”
transfers
ARIN sta's'cs
do not provide
any data about
the iden'ty of
transferors
and transferees.
?
ARIN
?

| 7
A.2) IPv4 Address Transfer Recipients: 2014/2015
Total: 38.9 million
Top 10: 83%
HHI >1980
Total: 10.2 million
Top 10: 61%
HHI > 1470
Total: 18.3 million
Top 10: 42%
HHI > 260
ARIN RIPEAPNIC
Source: : variations of delegated-arin-extended
RIPE does
not report
on
“legacy”
transfers
Reversed
Engineered
?

| 8
A.3) IN-Region vs OUT-of-Region
2014-01-01 to 2015-12-31 Source: ARIN, APNIC, RIPE: APNIC & RIPE transfer stats files
Addresses are flowing from ARIN to the other regions.
ARIN
38.9 million
APNIC
10.2
million
RIPE
18.3 million
From 11/23/2015
to 12/31/2015

| 9
A.3) How “Old” Were ARIN Addresses Transferred
to APNIC? 2014-01-01 to 2015-12-31 2014-2015
0
500000
1000000
1500000
2000000
2500000
3000000
1986 1988 1989 1990 1991 1992 1993 1994 1995 1997 1998 1999 2001 2002 2006 2007 2008 2009 2012 2014 2015
Addresses Transferred By Previous Registra=on Date
Pre-ARIN ARIN Pre-ARIN via ARIN transfer
Legacy addresses
transferred to ARIN in
2014-2015 prior to
being re-transferred to
APNIC to a subsidiary
4%
Legacy
addresses
79%
ARIN
addresses
17%
ARIN
crea'on

| 10
A.4) Distribution of Address Block Size in Transfers
2014-01-01 to 2015-12-31
-
2,000,000
4,000,000
6,000,000
8,000,000
10,000,000
12,000,000
14,000,000
16,000,000
18,000,000
/8 /9 /10 /11 /12 /13 /14 /15 /16 /17 /18 /19 /20 /21 /22 /23 /24
Addresses Transferred per Size of Address Block Transferred
ARIN APNIC RIPE
Larger transfers Smaller transfers

| 11
A.5) Evolution Over Time
2014-01-01 to 2015-12-31
-
5,000,000
10,000,000
15,000,000
20,000,000
25,000,000
30,000,000
35,000,000
40,000,000
45,000,000
Cumula=ve Number of Transferred Addresses
ARIN APNIC RIPE

| 12
B) Growth of RIRs IPv4-Delegated Table Source: delegated files
Why is this important to track over time?
The global BGP table derives from the RIR table and contains about 600,000 entries
on Jan 1st 2016, roughly 4 times the number of entries in the RIR table (due to
factors including internal de-aggregation and traﬀic engineering). Any increase in
the RIR table could then create a significant surge in the BGP table.
ARIN.. APNIC.. RIPE.. LACNIC.. AFRINIC.. Total..
1/2012 43,739 19,806 44,130 3,714 1,926 113,315
1/2013 45,410 21,144 48,643 4,001 2,145 121,343
1/2014 52,047 22,742 50,004 7,800 2,382 134,975
1/2015 54,438 26,773 51,319 9,373 2,582 144,485
1/2016 56,852 31,616 56,105 10,798 2,857 158,228
4 Years later: +30% +60% +27% +191% +48% +40%
Counting increases of IPv4 assigned or allocated entries over the last 4 years

| 13
RIR Contributions to IPv4-Delegated Table Growth
Year over Year
0%
2%
4%
6%
8%
10%
12%
2012 2013 2014 2015
ARIN APNIC RIPE LACNIC AFRINIC
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
2012 2013 2014 2015
ARIN APNIC RIPE LACNIC AFRINIC
Relative Contribution Absolute Contribution
Source: delegated files

| 14
C) Need-Based Policies and Private Contracts
The eﬀect of need-based policies is to limit the size of the address
block being transferred to the actual assessed need of the recipient.

They do not prevent private contracts between par'es such as
Le[er of Authoriza=ons and Op=ons.

Such contracts and are not recorded publicly, thus it is impossible
to measure the number of IPv4 addresses under those contracts
and evaluate the concentration of this derivative market.

| 15
Note about Statistics Collection
ARIN, APNIC & RIPE report diﬀerent data about transfers.
This makes data analysis across regions diﬀicult.
Origin
Org
Id
Dest.
Org
Id
Original
Block
Transferred
Block
Previous
Registra=on
Date

New
Registra=on
Date

Country
of
Origin
Country
of
Des=na=on

Format
ARIN ✔ ✔ WEB
APNIC ✔ ✔ ✔ ✔ ✔ ✔ ✔ TXT
RIPE ✔ ✔ ✔ ¢ ✔ JSON
Note: ¢ RIPE does not report transfers of legacy blocks

IPv4 Transfer Analytic View

More Related Content

Similar to IPv4 Transfer Analytic View (20)

More from APNIC (20)

Recently uploaded (20)

IPv4 Transfer Analytic View