Review of Bedrock AI – a screening tool for risk

I was approached on Twitter if that they liked my tweets (at least one 🙂 ) and asked if I wanted a free trial of the product. I said yes and tried it …

First impression: A very neat and modern design.

First Test: Checkpoint is my biggest position and I deem it a high quality company, especially the management/governance aspect. I view Tesla as rather the opposite. So, I did a quick test, which supported my notion. Current risk scores were 26 for CHKP (down from 35/39 for 2018/19) vs 75 for TSLA (worse than 2018/19 at 46/71).

Progressive is another higher quality company (as far I am concerned, influenced by others) that I watched next. PGR, as well, has a low risk score of 34. As an insurance company, it does not surprise that “Top Red Flag Categories” are accounting heavy. Peers are in the 48 (AIG) to 61 (ALL) range.

Alibaba’s 80 is no surprise either.

ZIM has a 47 risk score. Top Red Flag Categories are (1) Related parties, conflicts & other relationships (2) Governance and (3) Accounting policy. And (1) related parties are indeed a huge issue here!

  • Kenon Holdings LTd. is a major shareholder (27.8%) and ZIM has multiple transactions (incl. Chartering in vessels from) with related parties, which are connected to Kenon
  • 8 (out of c. 90) vessels are chartered in from a related party

Tencent Music (TME) worsened to 82 from 62 (20-F, 2018). Top cateogires are

  • Related parties, conflicts & other relationships
  • Governance
  • Subpoenas, investigations & other infractions

Some more

  • ATUS: 71, constantly worsened (more flags) from Dec 2018, with 10-Ks not deviating from trend
  • TEVA: 65 (10-Q, Sept’21) like last quarters, last three 10-Ks are worse (82, 81, 75)
    • Top category: Legal & regulatory
  • QRTEA: 54 much improved since 10-Q, Sept’20 (72), 10-K 2018 was 80
  • PBR: surprisingly good with a 40 score, improved from 20-F Dec’19 (41) and 20-F Dec’18 (70), but the first flag reads “Operations with related parties may not be properly identified and handled”
  • BBBY has a 61 score (10-Q, Nov’21), but was in a much higher range (81-83) earlier (10-K Feb’20 until 10-Q Nov’20). Top catories are (1) Impairment, (2) Accounting policy and (3) Buzzwords & jargon.

Risk scores seem correct or at the very least they go into the right direction on average … or my impressions and right feelings could be grossly wrong! But there is also …

Some proof. Bedrock published a White Paper supporting a thesis of higher returns through excluding high risk filers (or shorting these). Out-performance is more consistent for longer periods (12m+).

I find Bedrock AI a useful tool, without claiming to fully understand the methodology or having figured out how best to use it (yet). So far, I have mostly looked at the total risk score, ist development over time and top categories. I have clicked on some flags and read though some text snippets from the SEC filings but have not gone further.

I recommend to try Bedrock AI. You can sign up here* for a trial. Eligible sign-ups will receive an extra week free if you mention “Searching4Value” in your message.
(* I do not receive anything for this)

Do you use any such helpful tools ?


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s