Here's a simplistic statistics of results using various anti-virus products against 2614 malware collected by nepenthes.
Antivir: 2327 / 89.02 %
AVG: 2309 / 88.33 %
ClamAV: 1890 / 72.30 %
F-Prot : 1890 / 72.30 %
Total number of binaries that were not identified by all of the antivirus products: 201. Alternatively, you can view the
statistics by the nepenthes team:
Rank Product Hit Rate Trend
1 Antivir 99,04% +7,07%
2 BitDefender 96,23% +1,52%
3 VirusBlokAda 95,17% +1,42%
4 F-Prot 94,02% +2,39%
4 Authentium 94,02% new
5 Norman Virus Control 93,78% +1,19%
6 Fortinet 87,29% +2,35%
7 F-Secure Antivirus 85,22% +5,99%
8 Kaspersky 85,10% +5,73%
9 VirusBuster 82,53% +11,76%
10 Trend Micro 76,19% +5,14%
11 ClamAV 71,41% -0,85%
12 NOD32 70,06% +4,05%
13 Sophos SWEEP 68,58% +2,45%
14 eTrust 63,97% new
This
paper, titled "Automated Classification and Analysis of Internet Malware" is a good read describing the limitations and inconsistencies of anti-virus malware detection across several AV products. It offers a good solution: describing and classifying malwares based on their behaviors (hence, behavorial analysis).
Speaking of behavorial analysis, I would like to point out that this is one of the two long term projects that the Malaysia Honeynet Project plans to undertake. The other project is deployment of high interaction honeypots. These two projects however, requires sponsorships, specifically hardware, bandwidth and network. I will write more about this.