[philrapo]

Prominently displayed in large font on page 1 of the Goldman research report on Tesla:

"Goldman Sachs does and seeks to do business with companies covered in its research reports. As a result, investors should be aware that the firm may have a conflict of interest that could affect the objectivity of this report. Investors should consider this report as only a single factor in making their investment decision."

(I put this in the reply to a child, but I felt it was relevant enough to reply directly to the parent.)

[btown]

If you want to evaluate the quality of Goldman's research analysis on its own merits, this blog post picks out the core of the valuation argument (from mere hours before the IPO announcement)... http://www.zerohedge.com/news/2016-05-18/goldman-compares-te...

The only way they get from a $125 base case to the final $250 valuation is by assigning a disjoint probability to the idea that even if Musk doesn't build Steve Jobs-level hype, he would build Maytag Repairman-level hype around electric vehicles; therefore, they can double the probability that there is an above-fundamentals scenario. Whether or not a wall was broken, this is kind of ridiculous.

(And, regarding walls: http://www.cnbc.com/id/100774459)

[ryguytilidie]

Do you think that putting a disclaimer of your sketchy behavior excuses your sketchy behavior?

[hueving]

That's par for the finance world. They will sell you shit with a smile as long as it pays commission even though they would never touch it (or are even short it).

[martin1975]

Kind of like "collateralized debt obligation", a.k.a. shit wrapped in shit.... circa 2008.

[spyspy]

For anyone whose knowledge of this starts and ends with The Big Short, CDOs are not inherently bad. They're just a structure for divvying up assets based on risk. The problem up to 2008 was that the risks were downplayed/hidden by the banks and the credit rating agencies.

[eru]

We do the equivalent of CDO to our networks at Google. (Ie tcp can't deal with packet loss, so needs the best tranche; especially if user facing. Some long running copy operations with fountain codes only care about total throughput, not lost packages, so get the `equity' tranches.)

[cjarrett]

If someone is thinking citation needed: http://www.sec.gov/spotlight/enf-actions-fc.shtml

[monochromatic]

Aren't they by definition short any security they're selling you?

[TrickedOut]

No - when they sell you a security, the assumption is they were long until then (holding it) and then either disposed of their entire long position or disposed part of it (in which case they continue holding it long.)

The case where they are selling you something and are SHORT the product is a special case which requires more work usually. The most egregious case was the ABACUS deal (http://www.theworkingeconomy.com/simple-explanations-of-econ...) where Goldman was actively short the product and needed a dumb customer to unload the other side (enter Fabulous Fab and his widows orphans: http://www.reuters.com/article/us-goldman-emails-idUSTRE63O2...)

[hueving]

No, not if they are just acting as a broker.

[pdabbadabba]

Seems to me that depends on the actual quality of the report. If the report is substantively useful, and not hopelessly biased, then I don't see anything too sketchy about this.

[mattmaroon]

Anyone reading a Goldman research note is going to be savvy enough to be aware of the conflict.