March 1, 2009 marked the end of Greenbackd’s first quarter, so we thought we’d take the opportunity to update you on the performance of the Greenbackd Portfolio and the positions in the portfolio, discuss some changes in our valuation methodology since our first post and outline the future direction of Greenbackd.com.
First quarter performance of the Greenbackd Portfolio
We get many questions about the content and performance of the portfolio. We had originally planned to report on a six-monthly basis, but we have now decided to report on a quarterly basis so that we can address these questions on a more frequent basis. Although it is still too early to determine how Greenbackd’s strategy of investing in undervalued asset situations with a catalyst is performing, we’ve set out below a list of all the stocks we’ve included in the Greenbackd Portfolio and the absolute and relative performance of each at the close on the last trading day in our first quarter, Friday, February 28, 2009:
The absolute total return across the current and former positions as at February 28, 2009 was -3.7%, which was +7.0% higher than the S&P500’s return over the same periods. A negative return for the first period is disappointing, but we are heartened by the fact that we outperformed the market by a small margin.
You may have noticed something odd about our presentation of performance. The S&P500 index declined by 18.0% in our first quarter (from 896.24 to 735.09). Our -3.7% performance might suggest an outperformance over the S&P500 index of +14.3%. We calculate our performance on a slightly different basis, recording the level of the S&P500 index on the day each stock is added to the portfolio and then comparing the performance of each stock against the index for the same holding period. The Total Relative performance, therefore, is the average performance of each stock against the performance of the S&P500 index for the same periods. As we discussed above, the holding period for Greenbackd’s positions has been too short to provide any meaningful information about the likely performance of the strategy over the long term (2 to 5 years), but we believe that the strategy should outperform the market by a small margin.
Greenbackd’s valuation methodology
We started Greenbackd in an effort to extend our understanding of asset-based valuation described by Benjamin Graham in the 1934 Edition of Security Analysis. Through some great discussion with our readers, many of whom work in the fund management industry as experienced analysts or even managing members of hedge funds, we have had the opportunity to refine our process. We believe that what started out as a pretty unsophisticated application of Graham’s liquidation value methodology has evolved into a more realistic analysis of the balance sheet and the relationship of certain disclosures in the financial statements to asset value. We’re not yet ready to send it into space, but we believe our analyses are now qualitatively more robust than when we started and that has manifest itself quantitatively in better performance (more on this below).
The two main differences between our early analyses and our more recent ones are as follows (these are truly cringe-worthy, but that’s why we undertook the exercise):
- We didn’t take account of the effect of off-balance sheet arrangements and contractual obligations. This caused us to enter into several positions we should have avoided, including BGP and VVTV.
- We were using overly optimistic estimates for the recovery rates of assets in liquidation. For example, we started using 50% of Gross PP&E. We now use 20% of Net PP&E. We now apply Graham’s formula as the base case and deviate only when we believe that Graham’s formulation doesn’t reflect reality.
The effect of these two broad errors in analysis was to create several “false positives,” which is to say that we added stocks to the portfolio that wouldn’t have passed our current, more rigorous standards. The performance of those “false positive” stocks has been almost uniformly negative, and dragged down the performance of the portfolio. As an exercise, we went back through all the positions we have opened since we started the site and applied our current criteria, which are more stringent and dour than our earlier standards. We found that we would not have opened positions in the following eight stocks:
- BRN (-13.1% on an absolute basis and +4.9% on a relative basis)
- BGP (-10.8% on an absolute basis and -21.6% on a relative basis)
- COBR (-17.1% on an absolute basis and +3.6% on a relative basis)
- HRT (-25.3% on an absolute basis and -9.7% on a relative basis)
- KONA (+87.8% on an absolute basis and +81.9% on a relative basis)
- MGAM (-24.2% on an absolute basis and -5.0% on a relative basis)
- VVTV (-25.0% on an absolute basis and -23.1% on a relative basis)
- ZLC (-72.0% on an absolute basis and -61.1% on a relative basis)
It seems we got lucky with KONA, but the performance of the balance of the stocks was wholly negative. The performance across all stocks listed above was -12.5% on an absolute basis and -3.9% on a relative basis. Excluding these eight stocks from our portfolio (i.e. treating the portfolio as if we had not entered into these positions) would have resulted in a slightly positive absolute return of +0.7% and a relative performance over the S&P500 of +12.5%. This is a compelling reason to apply the more dour and rigorous standards.
We like to think we’ve now learned out lesson and the more dour and rigorous standards are here to stay. Set out below is an example balance sheet summary (for Chicago Rivet & Machine Co. (AMEX:CVR)) showing our present base case discounts from book value (circled in red):
Readers will note that these are the same base case discounts from book value suggested by Benjamin Graham in the 1934 Edition of Security Analysis, more fully described in our Valuing long-term and fixed assets post under the heading “Graham’s approach to valuing long-term and fixed assets.” Why we ever deviated from these standards in the first place is beyond us.
Update on the holdings in the Greenbackd Portfolio
Leading on from our discussion above, four of the stocks we picked using the initial, overly optimistic criteria no longer meet our more stringent standards but haven’t yet been removed from the portfolio. We’re going to take our medicine now and do just that. To make it clear, these stocks aren’t being removed because the value has deteriorated, but because we made a mistake adding them to the portfolio in the first place. As much as we’d like to treat these positions as void ab initio (“invalid from the beginning”), we’re not going to do that. We’ve made a full accounting of the impact they’ve had on the portfolio in the First quarter performance of the Greenbackd Portfolio section above, but we don’t want them affecting our future performance. The stocks to be removed from the Greenbackd Portfolio and their absolute and relative returns are as follows:
- BRN (-13.1% on an absolute basis and +4.9% on a relative basis)
- HRT (-25.3% on an absolute basis and -9.7% on a relative basis)
- MGAM (-24.2% on an absolute basis and -5.0% on a relative basis)
- COBR (-17.1% on an absolute basis and +3.6% on a relative basis)
We’ll provide a more full discussion of where we went wrong with these stocks at a later date, but suffice it to say for present purposes that all were errors from the second bullet point in the Greenbackd’s valuation methodology section above (i.e. overly optimistic estimates for the recovery rates of assets in liquidation).
There are fifteen stocks remaining in the Greenbackd Portfolio:
Eight of these positions (ABTL, ACLS, ARCW, CAPS, CRC, CRGN, NSTR, and VOXX) are trading at or below our nominal purchase price and initial valuations. The remaining seven positions (AVGN, DITC, IKAN, MATH, NENG, NTII, and SOAP) are trading above our intial purchase price but are still at varying discounts to our valuations. We’ll provide a more full update on these positions over the course of this week.
The future of Greenbackd.com
We are going to trial some small changes to the layout of the site over the next few weeks. We’ve already made the first change: the newest comments now appear at the top of the list. We’ll also be amalgamating some pages and adding some new ones, including a page dedicated to tracking the portfolio with links to the analyses. We’re also considering some options for generating income from the site. At the moment, Greenbackd is a labor of love. We try to create new content every week day, and to get the stock analyses up just after midnight Eastern Standard Time, so that they’re available before the markets open the following day. More than 80% of the stocks that are currently trading at a premium to the price at which we originally identified them (NTII, SOAP, IKAN, DITC, NENG, MATH and AVGN) traded for a period at a discount to the price at which we identified them. This means that there are plenty of opportunities to trade on our ideas (not that we suggest you do that). If you find the ideas here compelling and you get some value from them, you can support our efforts by making a donation via PayPal.
We look forward to bringing you the best undervalued asset situations we can dig up in the next quarter.
[…] that rises 25.0% in a quarter. Our Q1 performance was -3.7% (see our first quarter performance here), which means that our total return since inception (assuming equal weighting in each quarter) is […]
LikeLike
[…] we discussed in our review of our first quarter, we started Greenbackd in an effort to extend our understanding of asset-based valuation described […]
LikeLike
[…] of Greenbackd will note a few changes to the website. As we foreshadowed in our earlier post, Greenbackd Portfolio Q1 performance and update, we’ve amalgamated the old Contact us and Tips pages into a single Contact us / Tips page. […]
LikeLike
Oh one other thought I had (BTW I’m just a newbie and not sure of “proper” liquidation tactics)…
I notice that you don’t factor in any costs related to winding down operations. I guess this will be negligible for larger companies and ones with a huge margin of safety. But I wonder if this is something that should be discounted for some of the smaller ones. Although total liabilities is fully accounted for, is there a possibility of severence, pension, and similar liabilities that can crop up during a liquidation?
I just bring up this thought because your mistake last year seems to have been the oversight of off-balance-sheet liabilities. Being a total newbie to liquidations and financial analysis in general, I wonder if there can be sizeable liabilities not mentioned anywhere in the financial statements. I suspect in Graham’s days we never had pension liabilities or environmental liabilities like the present.
LikeLike
Sivaram,
The answer is that costs related to winding down operations should be included to the extent that they are real costs. Severance, pension, and similar liabilities are all real costs, and so should be included (most notably termination payments to management). We don’t explicitly calculate them because we’re looking for situations where the margin of safety is so large that those costs are negligible relative to that upside.
G
LikeLike
I just started following your blog a few months ago and although I can’t contribute financially or anything, I just want to say that you guys are doing a great job so far. I like how you detail mistakes with your technique and future improvements. The learning process alone may be worth more than any returns right now.
The reverse chronological posting is kind of confusing but maybe it takes some time to get used to it…
LikeLike
Thanks, Sivaram. We’re glad you’re getting some benefit from the site.
Green
LikeLike
It would also be interesting to see what the performance would be if the strategy is implemented with only stocks with a high probability of realization (a catalyst in advanced stages combined with steep discount to intrinsic value). My guess is that if you label each of your picks with a value that corresponds to a % confidence interval, your best picks would be those with a high probability of realization. For instance, you mentioned in one of your articles that SOAP was one of the best values you have found (probability of realization really high). It is not surprising to me that SOAP is in fact one of your best performers so far.
LikeLike
It’s an interesting thought, ef. We try to handicap the chance of realization from the outset, but not quantitatively i.e. not much further than a nominal characterization of “excellent,” “good,” “bad” etc. We’re prepared to accept lower upside if we think the chances of getting it are excellent (e.g. NSTR).
SOAP looked particularly good to us because it was one of the most deeply discounted net cash stocks we’d found (it was trading at about 40% of its net cash value) and management had taken steps to shut down part of the business so that its ongoing business was small in comparison to its net cash position. This meant it was unlikely to dissipate that cash quickly. We had no real idea about how effective Mark Nelson and Mithras would be, but we thought he had a *good* chance. That seems to be working out well, but I don’t think we could have predicted that at the start.
Contrast SOAP with AVGN, our best performer. AVGN was also deeply discounted (about 53% of net cash – not as deep as SOAP) but had a high historical cash burn rate, which meant there was a real risk that it could burn through its net cash value. We had no idea how effective BVF would be, but BVF has been an absolute revelation. Although its not yet finished, we think the campaign has been run superbly.
We’re not sure yet what the lesson is, but it certainly helps to have a simple situation (i.e. a stock trading at a discount to net cash).
How do you suggest we determine the chances of realization from the beginning? Is it a question of the stock or the catalyst (or some combination of both)?
LikeLike
Greenbackd,
Although it would be difficult to estimate the chance of realization quantitatively, I think they are a combination of both the catalyst and the margin of safety. For example, a stock trading below net-cash value even after a liquidation announcement certainly has a higher probability of realization than a company trading below net-cash but with no announcement made regarding liquidation or strategic alternatives.
I know it sounds intuitive, but I am mentioning this because Ben Graham developed a formula for arbitrage and special situations that takes into account the chance of realization. A portion in Security Analysis book appendix (1950s edition) actually discusses the use of the formula.
Ben Graham’s risk-arbitrage formula to determine optimal risk/reward is:
Return= CG-L(100%-C)/YP
Where:
• C is the expected chance of success (%).
• P is the current price of the security.
• L is the expected loss in the event of a failure (usually original price).
• Y is the expected holding time in years (usually the time until the merger or liquidation takes place).
• G is the expected gain in the event of a success (usually takeover price o liquidation value).
LikeLike