Talk:Net capital rule

July 2009 question and edits
Footnote 1 for this article says, " As is explained below, this is because investment bank holding companies were never subject to the net capital rule. Even the broker-dealer subsidiaries of these holding companies were not subject to a debt to equity leverage limitation. Instead, they were restricted to holding customer receivables not greater than 50 times the amount of their net capital."

This footnote appears to need a footnote. "As explained below"? Where, specifically, and on what authority? Steve.Murgaski (talk) 18:35, 9 July 2009 (UTC)

Sorry. The first sentence of footnote 1 is supported by the materials in footnote 12. The last two sentences are supported by the last five paragraphs of part 3 of the article (i.e the description of the Alternative Method) and more specifically by the materials in footnotes 29 and 30. I'd be happy to put those references into footnote 1 if you think that would be useful.

I just saw your comment and the changes someone else made to the article yesterday about "political content." I don't know what that would be. I realize the article differs from what a lot of people think the net capital rule is about, but I don't think anything in the article about the rule or its application to the large investment banks is inaccurate. I think before 2008 it would have been a dreary standard description, as indicated by the many footnotes to standard treatises and internet available GAO and CRS reports. The rule didn't change in 2008, but some new perceptions about it did develop. Please let me know if there is anything wrong in the article. The whole point of the article is to bring to Wikipedia the actual net capital rule described in long standing sources, not the popular perception presented in 2008 and 2009 materials that are quite at odds with the rule and its history and that are contradicted by the financial reports of the firms discussed in those materials.

EAFAAT (talk) 01:54, 27 July 2009 (UTC)

I have revised footnote 1 to explain what later text of the article supports the two points. I also added two “secondary sources” from before 2008 that show at least some people had recognized earlier that investment bank leverage in the 1990s exceeded 15 to 1. The 1992 GAO report in footnote 28 already demonstrated such “secondary source” recognition of broker-dealer leverage. I also tried to correct the footnote cross references. I thought I'd corrected those last time, but I guess I didn't. The new footnote created another renumbering level.

I removed the “political warning” labels in the introduction and Section 1, because I do not understand to what they refer. Is there text in the introduction or Section 1 that is “political” or “ideological”? Is there something not factual in that text or elsewhere in the article? If anyone has any questions or comments on the article and would prefer not to post those questions or comments on this public discussion page, please contact me directly. I have been contacted by email that way, so I gather it is possible through my user name.

EAFAAT (talk) 20:51, 27 July 2009 (UTC)

Section 1 quotations
I could not understand the purpose of moving extended quotations from a footnote that used to be in Section 1. I have returned those citations to the footnotes in Section 1.1 and summarized the quotes. If anyone thinks it better to retain the full quotes in Section 1.1please do so.

Subheadings
In trying to understand the purposes of the August edits to the article, it occurred to me it would help to have more subheadings to reinforce the article’s narration. That’s why I added the detailed subheadings. The expanded Section 1 in particular seemed to need subheadings. Subheadings may also help avoid questions like the July 9 question on this discussion page.

Section 3.2
I elaborated on the difference between “leverage” and the Basic Method (particularly how it excludes secured debt) because this is missing in post-2007 commentaries. It was hard to find a FOCUS Report that used the Basic Method. I found that Seattle-Northwest used it until 2009. The fact only 1% of its liabilities qualified as “aggregate indebtedness” under the Basic Method on June 30, 2007 (as described in note 25 of this article), should be instructive to the many observers who talk or write about the net capital rule without knowing much, if anything, about it. For a recent example see page 3 of this October 29, 2009, Testimony to House Financial Services Committee.

Section 3.5 and the two methods throughout Section 3
I elaborated on the differences between the Basic and Alternative methods because there still seems to be so much disbelief that the CSE Brokers were not subject to a 12-1 or 15-1 leverage ratio. If the OIG CSE Report had confirmed the CSE Brokers all used the Alternative Method much of this confusion might have been avoided. Unfortunately, by repeating several times the textbook explanation that each broker-dealer could select between the two methods, that Report lost an opportunity to explain this was not relevant to the CSE Brokers. The detail I added should make it clear why it was a “no brainer” for the CSE firms to use the Alternative Method. If anyone thinks it’s too repetitive, please edit. The important thing to understand is that, while thousands of small brokers use the Basic Method, no large broker uses the Basic Method. It would make no sense for such a broker to use the Basic Method.

Section 4 introduction language
I revised the language about “exemptions” because there is much confusion on this topic. When they were careful, the authors of the OIG CSE Report and the of the SEC proposing and adopting releases for the 2004 rule change stated the “exemption” was from the standard “haircut” method for computing net capital. Unfortunately, the unqualified use of the word “exemption” has led to many statements that the CSE Brokers were exempt from the net capital rule. That seems to have led to the common statements that the SEC “repealed” the net capital rule or “released” the large investment banks from the rule. See the Wikipedia Henry Paulson entry for both formulations.

I added Bailout Nation as the source for the statement the rule change was known as the Bear Stearns exemption. That book states the name was given in 2004 when the rule change was adopted. This seems curious, given that Bear was the last broker to receive permission to use the method. Bear received its approval on November 30, 2005, more than 19 months after the SEC approved the rule change.

I added the names of the two bank holding companies that received approval to use the new method. It is not correct that only the five large “independents” received approval, as noted later in Section 4.2. The SEC estimated 11 brokers could qualify. Based on SIA’s 2004-2005 Yearbook, the likely missing four would be CSFB, UBS, Deutsche, and BofA. Since none of those banks shows indifference to regulatory capital requirements, the fact they didn’t apply for the “exemption” is another indication the “momentous” rule change wasn’t so momentous. A BofA FOCUS Report is referenced in note 30. It shows BofA Securities had tentative net capital of nearly $10 billion in June 2007, so it presumably could have applied for the “exemption.” The same FOCUS Report also shows GAAP leverage of about 64x. That was achieved while using the “traditional haircuts” under the net capital rule. As explained in note 30, the more appropriate comparison to holding company leverage would be 20x. Even that leverage is much higher than the mythical 15x or 12x limit that is so widely repeated.

I also revised the Hank Paulson “charge” language, because I could not find any support for the statement Paulson led the charge for the rule change other than an unsourced statement in a book written by the editor who made this addition to the article. Footnote 40 describes the sources I found. In footnote 40 I also referenced the SEC comment file on the proposed rule because it provides a record of who submitted comments and what SEC meetings with firms discussed the rule change (at least as noted by SEC personnel). I don’t propose this “proves” Paulson didn’t “lead” whatever “charge” there was, but it is a source for information rather than speculation. My own speculation is it’s unlikely an investment banker like Paulson knew much about the net capital rule. I also doubt he or Goldman thought it was the big deal that commentators decided it was in 2008.

Section 4.2
I highlighted the fact CSE Brokers only began using the new capital computation method in 2005. What is now note 50 had only shown the SEC approval dates. The point is important for a recent draft Working Paper written by Taylor D. Nadauld and Shane M. Sherlund titled “The Role of the Securitization Process in the Expansion of Subprime Credit”, which studies the effects of the “external shock” from the net capital rule change as part of a broader study whether (or thesis that) the search for collateral for AAA RMBS affected subprime mortgage underwriting standards by encouraging loan originations in areas of high house price appreciation. The Working Paper (on page 22) states that CSE Holding Companies increased their relative share of the securitized subprime loan market “beginning in 2003” and continuing “until 2005.” The authors were apparently unaware that meant the increased activity took place before Bear and Lehman (the two CSE firms most active in the subprime market) began using the new net capital computation method. Morgan Stanley (third among CSE firms for total subprime deals in Table 8 of the Working Paper) also only began using the method at that time.

Presumably because they thought the period from at least October 2003 (when the SEC first proposed a rule change) to 2005 would be affected by the rule change, the Working Paper’s authors failed to note (as shown on Table 8 of the Working Paper) the fastest growth in relative share for CSE firms took place from 2002 through 2004 (from 14.7% for 2001, up more than 50% to 22.9% in 2002, to 31.9% in 2003 and 43% in 2004) before any of the CSE firms used the new net capital computation method. Instead, the Working Paper notes (on page 22) the increase from 22.9% to 47.7% in the years 2003-2005. In 2006 the CSE firm share fell from 47.7% to 41.4%.

Since Bear and Lehman only began using the new capital computation method on December 1, 2005, Professor Coffee’s reasoning in note 6 of this article would lead to the conclusion the net capital rule change caused those CSE Brokers to reduce their subprime mortgage activities. I don’t propose any such causation. The Lehman 2006 10-K (available at the link in this article’s footnote 55) aptly states on page 14 (in the Item 1A Risk Factors) that in 2006 there had been a “downturn” in residential real estate values and that this threatened continued reductions in Lehman’s residential mortgage securitization activities, which (as noted on page 28) had “strong, but lower revenues” in 2006.

In fairness to the Working Paper’s authors they proceed from the assumption the 2004 rule change was a momentous “external shock” and do not purport to examine the rule change or its implementation. At page 20 they mischaracterize the net capital rule as a 2% net worth to asset requirement. Throughout the Working Paper they assume the net capital rule operates like a commercial bank leverage test. This may explain why they think the October 2003 announcement of a proposed rule change could have affected subprime securitization behavior from 2003-2005. That assumption is defeated by the fact (presumably unkown to the Working Paper’s authors) that the daily liquidation computation of net capital for Bear and Lehman (as well as Morgan Stanley) did not change until the end of 2005 and that even Goldman and Merrill (two smaller participants in the subprime market) did not begin using the method until after 2004.

Bank leverage ratios are computed quarterly and only reported a month after quarter end. Despite much misinformation about SFAS 157 (which misinformation should reference SFAS 115), most bank assets are still reported on a historic cost (hold-to-maturity) basis, adjusted only slowly through loss reserves. Broker net capital is computed daily based on current market values. It is a self-liquidation test, subject to “early warning” requirements to ensure its minimum liquidation “cushion” is available every day. Bank capital tests, particularly raw leverage tests, are not intended to be liquidation tests.

Updated SEC web links
I updated the SEC EDGAR website links. The SEC dropped the “IDEA” script on August 19, so the old links have been stale for nearly 3 months.

Recent changes in net capital argument
I notice some commentators have moved to saying the old net capital rule didn’t limit leverage, but that the 2004 change permitted broker-dealers to buy “risky assets” like RMBS. The standard rule has very crude haircuts for any investment grade rated debt, much like the crude “risk-buckets” in Basel I. Under those “haircuts” a BBB RMBS would get the same “haircut” as the AAA piece. I don’t think anyone would argue that is a good measurement of risk. In any case, BDs operating under the “standard” haircuts both before and after the 2004 rule change owned RMBS and corporate securities subject to the same 15c3-1(c)(2)(vi)(F) haircuts for investment grade debt securities.

Footnote 59 tries to address the issue of capital calculations under the two approaches. Someone should do some research on that issue.

Charlie Gasparino’s new book “The Sellout” asserts (at 197) the new VaR approach (or what he describes as any risk management tool the BDs chose) permitted brokers to treat any AAA rated security “basically the same” as government securities. The market didn’t price those two different types of securities the same. VaR would presumably value the capital requirement for RMBS based on historic volatility, so AAA RMSB (or other AAA securities) would only receive the same VaR charge as governments if their historic price volatility matched treasuries. In any case, the established “haircuts” also treated all investment grade rated debt securities “very much” the same as government securities. The difference in their haircut “value” was generally only 3% (i.e they had roughly 97% as much value for net capital as government securities). I wonder how long VaR actually yielded significantly lower haircuts. I assume that by early 2007 (and certainly by August) those VaR calculations for RMBS (at least subprime) were “off the charts.” Weren’t they already off the charts in late 2006 or even earlier? Does anyone have information about this?

The LBI 2002 10-K (available at the link in note 68 of this article) shows (at page F-29) that, long before LBI began using VaR to compute net capital haircuts, LBI and its subsidiaries were able to hold large RMBS positions ($9 billion) and (on pages F-26) were able to conduct large mortgage securitization operations. Page F-29 shows their total mortgage and corporate debt holdings roughly equaled their government and agency holdings. The FOCUS reports for Seattle-Northwest and BofA Securities in notes 25 and 30 show a disparity in holdings of “corporate” securities (which, in item 7D of the list of assets, would include RMBS and other ABS) and government securities (item 7B). The small broker (Seattle Northwest) holds far more government than corporate securities. This is presumably the picture Gasparino imagines was typical of brokers before the 2004 rule change. BofA Securities, however, holds nearly the same amounts of the two types of securities. As in the case of the LBI 2002 10-K, BofA accomplished this using the traditional haircuts. Perhaps it is comforting to Wall Street veterans to believe this difference between small and large BDs only developed after (and as a result of) the 2004 rule change that “doomed Wall Street.”

EAFAAT (talk) 02:23, 13 November 2009 (UTC)
 * Interesting. I found the article to be rather uninformed, biased, and slightly contradictory . Someone with expertise in the area needs to look it over. Unfortunately it's an abstract area and gets little attention.— DMCer ™  14:22, 4 January 2010 (UTC)

Removed chunk of info from Donaldson article.
I removed this] from William H. Donaldson's bio for obvious reasons. While it has more to do with this article than any, keep in mind that most of the content below is already discussed in this article. Also note that the content below was tagged with WP:UNDUE, WP:NPOV, and Template:BLP dispute, so I advise against pasting any large chunks into this article.— DMCer ™  15:19, 4 January 2010 (UTC)

February 23, 2010, edit
I’ve tried to retain the ideas added to the introduction in edits since 11/13/09 while making the introduction consistent with the rest of the article. Three sentences in the second paragraph of the introduction contained no sources and were not supported by the rest of the article. Footnote 3 was not relevant to that language. Although retained from the 11//13/09 version, it supported language in that version that had been deleted.

The net capital rule was not abolished for the 5 firms, as the cited sources in the article make clear. If one wants to argue it was “effectively abolished” that would need to be explained or at least stated by a secondary source. The threshold to apply for the new method was not $5 billion in assets. That would have permitted many broker-dealers to qualify. The BDs with $5 billion in tentative net capital had hundreds of billions in assets. If anyone believes it important to separately present the common error that the exemption was available to BDs with $5 billion in assets, please add that. I haven’t found the error in any pre-2008 sources, and I don’t think any smaller BDs applied for the exemption, so it would be misleading to suggest there was any ambiguity in the rule. Presumably, the article should include the correct threshold.

The belief that the SEC’s actions were “antithetical” to the purposes of the net capital rule is a judgment I’m sure many share. If a source is provided for that judgment in order to include it in the article, I suggest the SEC’s response be included (i.e. that the new $5 billion early warning requirement mitigated the reductions in haircuts and that in practice the new method worked, both during the crisis and since). That response is in the Sirri Speech and the SEC letter attached at pages 117-120 of the GAO 2009 Financial Crisis Report.

The SEC’s description of the purpose of the net capital rule is in the article. For a less formal description, see the GAO Risk-Based Capital Report at 130-131. A January 4 edit stated a possible interpretation of the SEC’s “premise” in enacting the rule (although the interpretation seems to have more to do with margin rules than the net capital rule), but it is not a summary of the article’s content nor a statement of a secondary source.

A January 4 edit referenced a 2004 no-action letter that “differs from the article.” Where that letter deals with a debt to equity requirement it is dealing with the paragraph (d) limit on subordinated debt serving as no more than 70% of a broker’s equity. I’ve added the point at the end of footnote 36. It didn’t seem important for the article. The sources cited in note 36 explain this “debt to equity” limit. I don’t believe anyone has claimed the net capital rule ever limited leverage to 3.3x (as a 30% equity requirement would imply). If anyone wants to see the primary source, paragraph (d) in the rule is (by the standards of the rule) quite short and clear. You can easily find it in the labyrinth of 15c3-1 by googling “SEC rule 15c3-1” and searching the rule’s text for “debt-equity requirements.”

If anyone thinks it important to include the subordinated debt limit in the article (not just in footnote 36), please add it. I don’t see how it is inconsistent with anything in the article. Was there something else in the no-action letter that the January 4 editor (or anyone else) thought inconsistent with the article? You’ll see from the SEC letter in the GAO 2009 Financial Crisis Report that the author of the no-action letter wrote the SEC’s letter commenting on the report. That the letter is very consistent with the article (as now revised so that the introduction is consistent with the body of the article).

I’ve tried to insert additional GAO and other secondary source references to answer the posted concern that the article may contain original research. When I originally worked on the article I used established treatises, law review articles, and GAO reports to show the actual nature of the rule. I thought multiple sources would demonstrate there are many long established secondary sources that provide that description, although it differs from the recent descriptions in newspaper and magazine articles. With the GAO 2009 Financial Crisis Report, it’s possible the bulk of the article could now be sourced to pages 130-153 of the GAO Risk-Based Capital Report and pages 36 to 42 and 117 to 120 of the GAO 2009 Financial Crisis Report. Please provide details of what information in the article appears to be without secondary source support, so that I can insert such sources. I’ll be away until March 8, but will check after I return.EAFAAT (talk) 18:49, 23 February 2010 (UTC)

Introduction language: "independent broker-dealers."
"For the five 'independent' broker-dealers the SEC exempted from the traditional haircuts (i.e. Bear Stearns, Goldman Sachs, Lehman Brothers, Merrill..."

I am curious about the use of "independent" here, since, in the business, an independent broker-dealer is almost the opposite of what these giant firms are: the term usually means a firm is solely engaged in broker-dealer business. Not investment banking or any other financial services activities and not a subsidiary of a diversified holding company.

Can we defend the article's odd-looking use of the word "independent?" I realize the probative value is technically limited, but look at en.wikipedia.org/wiki/Broker-dealer to confirm my version of the definition. HedgeFundBob (talk) 03:20, 25 March 2010 (UTC)

Thanks for pointing this out. I had used "independent" in the way Professor Coffee did in his article cited in fn. 26. It really just meant an investment bank that was not part of a larger commercial banking organization or "bank holding company" as I've revised it. The only significance is that, despite the often repeated misstatement that only five large BDs were exempted from the traditional haircuts, the SEC granted the exemption to two BHCs and their BDS (i.e. to Citigroup and JP Morgan Chase). There have been edits to this article to repeat the misstatement that only 5 BDs were exempted from the traditional haircut, so I didn't want to leave the "five" language unqualified.EAFAAT (talk) 20:07, 27 March 2010 (UTC)

May 26, 2010, edit
I’ve added information from the FCIC hearings and recent books and more technical papers. I shortened the introduction because it has expanded through comments that relate to the text of the article. If anyone thinks it better to keep in the introduction the NY Sun example of different haircuts for Treasuries and equities, please put it back there. I moved it into the text where haircuts are discussed. I think it’s fine to refer to haircuts as covering market risk, but I’ve inserted the standard wording “risk characteristics” from the GAO because that standard wording distinguishes among market, volatility, and liquidity risk.

I’m not sure about Section 7.1. I thought it relevant to include Merrill’s statement in its 2005 10-K, but I couldn’t find anything indicating the significance of other net capital fluctuations. I’ve just put in raw data without analysis. I have all the 10-Q and 10-K reported net capital levels from 2003-2009 but figured that was too much information.

I added the 10-Q leverage data from the 1990s because it occurred to me limiting leverage discussion to 10-Ks could indicate there was otherwise more post-2004 leverage. The FRBNY paper gives a useful chart of the much higher 1990’s leverage.

In my November 13, 2009, comments I was wrong about Gasparino’s book. He doesn’t really write that BDs held treasuries before 2004 (although I’ve heard him say that). He actually details in the book various increases in MBS holdings by investment banks during the 1990s and otherwise before 2004. He also notes several times their very high pre-2004 leverage, particularly in the 1990s.EAFAAT (talk) 02:36, 26 May 2010 (UTC)

June 2010 edits
The language preceding footnote 89 is not a complete sentence, so I thought I needed to correct that change made on June 26. The reason I originally footnoted in the middle of the sentence was to show the footnote explained Goldman did not disclose net capital levels (so only 3 of the 4 remaining CSEs did). Maybe footnotes 89 and 90 should be combined so that the entire paragraph is covered by a single footnote. The other June 26 change to "the" 2004 rule change seemed fine. I originally wrote "a" because I thought the section was the beginning, with the introduction just summarizing what came below.EAFAAT (talk) 17:03, 28 June 2010 (UTC)

February 2, 2011, edit
I removed ibid from the footnotes, so I deleted from the article the request for that change. I inserted the FCIC Report’s statements about the net capital rule change and the CSE program. They are consistent with the article, although some relevant items in the article are not mentioned in the Report. I included Professor Hall’s speech and the snafu in the FCIC Report’s quotation of Sirri, because both indicate how easy it is to make mistakes in this area.EAFAAT (talk) 00:20, 2 February 2011 (UTC)

March 10, 2014 - Paragraph Layout
The discussion of the impact the Net Capital Rule has on the 2007/8 banking collapse doesn't belong as the first section. You lead with it's history, and then you move on (chronologically) from there. Even if that section should not be in the chronology of the rule, (I think it _probably_ shouldn't, but it's a judgement call) then I'd say it stands as a separate section, but still AFTER the chronology.

Talk about what the rule is. Then talk about how it's changed. Then get to the stuff that may be controversial. — Preceding unsigned comment added by 170.20.11.28 (talk) 00:08, 11 March 2014 (UTC)

Risk-weightings post 2004
I hope EAFAAT can answer a few questions about his edit. Our article says:
 * "In 2004 the SEC amended the net capital rule to permit broker-dealers with at least $5 billion in "tentative net capital" to apply for an "exemption" from the established method for computing "haircuts" and to compute their net capital by using historic data based mathematical models and scenario testing authorized for commercial banks by the "Basel Standards.""

and:
 * "The SEC expected this change to significantly increase the amount of net capital computed by those broker-dealers."

Further down it says:
 * "Studies of Form 10-K and Form 10-Q Report filings by CSE Holding Companies have shown their overall reported year-end leverage increased during the period from 2003 through 2007 ..."

1. There are constant comparisons with post-2004 leverage and pre-2004 leverage even though the computations are made with different risk-weighting (haircuts). This is an apples and oranges comparison that doesn't help the reader understand a quantitative comparison.

2. The article refutes the notion that the 2004 changes in haircuts brought about an immediate change in risk-taking. Yet, it also says leverage increased from 2003 to 2007. Was it just a slow change?

3. I don't see the article addressing critic's claims that unweighted risk increased. One way to do this would be to used the same haircuts to compute leverage in both pre-2004 and post-2004. Both Ed Conard in his book "Unintended Consequences" and Friedman/Kraus in "Engineering the Financial Crisis" specifically cite the change to risk-based weightings in increasing raw un-weighted leverage numbers and devote pages if not a chapter to this idea. If this is wrong, the article should explain. Jason from nyc (talk) 18:35, 13 March 2017 (UTC)

External links modified (February 2018)
Hello fellow Wikipedians,

I have just modified 2 external links on Net capital rule. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://web.archive.org/web/20090831071357/http://treas.gov/press/releases/reports/hedgfund.pdf to http://www.treas.gov/press/releases/reports/hedgfund.pdf
 * Added archive https://web.archive.org/web/20110130170912/http://c0182732.cdn1.cloudfiles.rackspacecloud.com/fcic_final_report_full.pdf to http://c0182732.cdn1.cloudfiles.rackspacecloud.com/fcic_final_report_full.pdf

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 08:33, 16 February 2018 (UTC)