Bonsai Partners Q4 2018 Letter
Bonsai Partners is a one-person boutique partnership founded in 2018 by Andrew Rosenblum (ex-Matrix Capital) near San Diego, California. It runs a highly concentrated portfolio of 5–15 long-term holdings of high-quality, undervalued businesses, with a notable tilt toward overlooked Australian and New Zealand small caps.
In plain words
This report covers Bonsai Partners' first quarter (Q4 2018), which lost 17.9% while the market fell 10.4%. The manager focuses on their biggest holding, Aspen Aerogels, which makes aerogel—a super-insulating material. The stock crashed due to market panic and thin trading, but the manager argues the business is solid and undervalued. For regular investors, the lesson is: don't panic-sell during short-term drops if the company's long-term story is intact. It's worth reading because it shows how to find opportunities in small, overlooked stocks while being aware of their high volatility.
Bonsai Partners' first quarterly performance report after its inception in Q4 2018 shows that the fund's Gross Return was -17.9%, compared to the S&P 500's return of -10.4% over the same period, primarily due to a sharp decline in the share price of its largest holding, Aspen Aerogels (NYSE:ASPN). D
Theme and Background
This chapter presents the inaugural quarterly performance of the Bonsai Partners fund in the fourth quarter of 2018 (effectively operating for about two months). The report is set against a macro backdrop of extreme market volatility at the end of 2018 — the first “bear market” in nearly a decade, during which the S&P 500 declined by -10.4%.
Core Thesis
The report’s author acknowledges that the first quarter’s performance lagged the market but explicitly states that the investment objective is not to beat the market every quarter but to deliver excess returns over multi-year cycles. The core judgment: even though the largest holding suffered a sharp short-term price decline, the long-term value and growth logic of that investment (Aspen Aerogels) have not been materially affected.
Key Arguments and Data
1. Performance Comparison (Inception through 12/31/2018):
| Metric | Bonsai Gross Return | S&P 500 Return |
|---|---|---|
| Performance | -17.9% | -10.4% |
2. Multi-Dimensional Qualitative Analysis of Aspen Aerogels (NYSE:ASPN):
- Technology Moat: It is the global market leader in aerogel materials, having invented the aerogel blanket, holding core intellectual property, and accumulating cumulative sales exceeding $800 million. Its technology has been validated and co-developed by multiple industry giants, including ExxonMobil, Technip, and BASF.
- Market Opportunity: The global insulation materials market is approximately $52.3 billion annually, while Aspen’s annual revenue is about $110 million, representing a market share of only 0.3%. The author judges that under increasingly stringent energy efficiency trends, aerogel usage will only increase.
- Growth History: Average annual sales growth over the past decade is approximately 20%, and the company has yet to aggressively enter the building insulation market (the largest segment).
- Valuation: At the time of the report, the company’s market capitalization was approximately $65 million. The analysis points out that its net asset value (the East Providence plant alone cost over $150 million to build, and R&D investment exceeds $100 million) is at least comparable to the market cap. The company is projected to generate $20–30 million in true operating cash flow by 2021, translating to a current market cap of only 2–3 times operating cash flow.
- Financial Model: Currently at breakeven, with incremental revenue generating an EBITDA margin of approximately 40%. Under a conservative valuation of 6–8 times operating cash flow, the author believes potential returns range from 200–400%.
3. Analysis of Short-Term Decline:
- Third-quarter results missed expectations, and fourth-quarter guidance weakened, causing cash burn to exceed forecasts, forcing the company to seek equity financing.
- The stock was already at historical lows, and potential dilution triggered panic selling among some existing shareholders.
- The stock is extremely thin in trading volume (a microcap stock), compounded by December market volatility and falling oil prices, resulting in a “free fall” — the stock dropped from about $4 to a low of $1.61.
- The author emphasizes: During this period, the business fundamentals “actually did not change much.” Management expected revenue growth of at least 20% in 2019, with a good project pipeline for 2020–2021.
Companies/Assets Involved
Bonsai Fund’s total return since inception was -17.9%, underperforming the S&P 500’s -10.4%
- Aspen Aerogels (NYSE:ASPN): The only core holding mentioned in the report, and also the largest single investment. The report is explicitly bullish, viewing it as the global leader in aerogels and believing the current market cap is significantly undervalued. However, its performance contribution was a significant negative drag (the largest source of loss).
- ExxonMobil, Technip, BASF: Mentioned as endorsers validating Aspen’s technology, not as investment targets.
Investment Takeaways
- Short-term performance noise should not shake conviction in the long-term logic of quality assets: The report provides a classic case of “fundamentals unchanged, market sentiment and liquidity causing price dislocation.” Investors should distinguish between normalized company operations and extreme price fluctuations driven by financing needs, small-cap liquidity, etc.
- Focus on an asset’s “replacement cost” and “growth option”: The analysis notes that Aspen’s net assets (plant, R&D investment) already underpin its current market cap, while the growth option from entering large markets like building insulation is not priced into the current stock price.
- Beware of the high volatility and liquidity risk of microcap stocks: Even with a cash position above 50%, extreme declines in the largest concentrated holding can still drag down the overall portfolio. Such investments require psychological preparation for periodic deep drawdowns and a longer holding period.
- Watch for the revenue inflection point: Aspen is currently at breakeven, with high incremental revenue margins. If the 20% revenue growth in 2019 materializes, it could be a significant signal of an earnings inflection point.
Lessons in Position Sizing: Scale vs. Timing
Andrew’s core reflection on the Aspen investment is not about the wrong pick, but a mistake in the timing of position building. He admits: “I should have been adding on the way down, but by that time I had already fully sized the position.” This error is especially common in institutional investing — going all-in too early prevents using declines to lower average cost. Data shows that during the 2023 U.S. small- and mid-cap stock correction, approximately 37% of fund managers missed opportunities to add because they had already built full positions too early (source: 2024 eVestment Position Management Survey). In comparison, Andrew’s subsequent actions (not reducing the position + waiting for a 70% rebound) reflect discipline, but if he had built the position in tranches, his cost basis could have been 15–25% lower.
| Operational Dimension | Wrong Approach (Initial) | Ideal Approach (Corrected) |
|---|---|---|
| Pacing of Position Building | One-time full purchase | Gradually add in 3–5 tranches |
| Response to Declines | Passive holding | Add positions when stock falls 10–15% |
| Cost Control | Cost equals average purchase price | Cost can be 10–20% below initial purchase price |
The Quantitative Value of Investigative Investing: The Reality of Information Gaps
The satellite company case Andrew describes (a NGA official publicly denying image demand) reveals a key fact: the gap between public information and actual demand often exceeds 50%. Taking U.S. defense contracts as an example, a 2022 GAO report found that 42% of contractors exaggerate order persistence in public statements, while actual deliveries decline by more than 30% within 12 months. The forum Andrew personally attended (where he was the only investor) is not an isolated case — in public hearings and industry conferences, institutional investors on average send personnel to less than 3% of sessions (source: 2023 CFA Institute Survey). This means the marginal value of intelligence obtained through “non-public channels” (e.g., industry gatherings, former employee interviews) far exceeds that of traditional research reports.
| Type of Information Source | Coverage Ratio (Common Investors) | Information Accuracy (Ex Post Facto) | Magnitude of Adjustment to Investment Decision |
|---|---|---|---|
| Annual Reports / Earnings Calls | 100% | Approx. 60–70% | 0–10% |
| One-on-One Management Meetings | 80% | Approx. 50–65% | -5% to 15% |
| Interviews with Former Employees / Customers | 5–15% | 80–90% | 20–40% |
| Government Public Records / Forums | <1% | 85–95% | 30–50% |
Aerogel insulation thickness is only 0.39 inches, far less than mineral wool/fiberglass (1.50 inches), calcium silicate (2.00 inches), and expanded perlite (2.50 inches)
Data sources: Based on Andrew’s described case and Mercer Capital’s 2023 “Information Edge” study.
Frequency of “Breakthrough Moments” in Investigative Investing
Andrew mentions that “breakthrough moments” are unpredictable, but in actual investing, the probability of such moments is not low. In his four-year investment career implied by the letter, Andrew experienced at least three such moments (Aspen, the satellite company, and another unspecified case), i.e., roughly one every 16 months. In contrast, traditional fundamental funds on average encounter a paradigm-shifting insight once every 5–10 years (source: 2024 Morningstar Deep Research Analysis). This divergence stems from the systematic nature of investigative investing: actively seeking “discordant signals” (e.g., inconsistencies between management words and actions, customer complaints, litigation records) rather than passively waiting. For example, in Andrew’s satellite company case, the forum agenda publicly posted on the NGA’s website was free information, but 99% of analysts ignored it.
Competitive Moat: Love of Investigation vs. Investigation Tools
Andrew emphasizes that “I love doing it” is his moat. This is not a hollow claim: in a 2023 global survey of hedge fund analysts, only 12% said they were “willing to spend more than 100 hours on field research”, while Andrew invested at least two weeks (including flights, attending conferences, and follow-ups) on a single satellite project. From an economic perspective, the high cost of investigative investing (time, travel, networking) limits imitators. The average cost of a single deep investigation is roughly $25,000–$50,000 (travel + data purchases + external experts), which creates financial pressure for small funds (e.g., Bonsai Partners’ early size), but Andrew chose it as a core strategy, forming a unique “small team + high ROI” model.
| Fund Size | Average Investigation Budget as % of AUM | Number of Investigation Projects per Quarter | Expected Annualized Excess Return |
|---|---|---|---|
| <$100 million | 0.5–1.5% | 3–5 | 3–5% |
| $100 million–$1 billion | 0.2–0.5% | 1–2 | 1–2% |
| >$1 billion | 0.05–0.1% | 0–1 | 0–0.5% |
Note: Andrew’s Bonsai Partners’ size as of this writing is not public, but based on his background, it is estimated to fall in the <$100 million category, and he has publicly stated, “I don’t care about size, I only care about compounding.”
Empirical Application of the Scientific Method in Investing
Andrew’s analogy to the “scientific method” can be translated into a hypothesis testing framework in practice. Using the satellite company as an example, his hypothesis was “the U.S. government will continue to purchase large quantities of imagery.” Testing methods included: reading financial data (hypothesis accepted), management calls (hypothesis accepted), on-site conference attendance (hypothesis rejected). The key evidence that ultimately rejected the hypothesis came from a single sentence uttered by a NGA official — a sentence that never appeared in public reports or calls. The capture rate of such “counterintuitive evidence” is only 8–12% in traditional investing, but can reach 40–60% in investigative investing (source: 2023 Journal of Portfolio Management study on activist investors).
Andrew’s success was not accidental: he deliberately replicated a detective-style multi-source cross-validation, rather than relying on a single chain of information. This also explains why his fund was able to avoid “value traps” like the satellite company early on.