COE explained: a structured note sold by a bank whose return depends on a reference asset, usually capped, usually carrying 4% to 7% in embedded commission. Most of it loses to Tesouro Direto, according to the distribution sources the video cites.
COE explained: what the acronym hides
COE explained plainly is a Certificado de Operações Estruturadas, a structured note issued by a bank whose return depends on a reference asset rather than a fixed rate. The transcript's argument is that the payoff formula, not the headline scenario, decides whether the investor keeps anything after fees.
The product is built by a bank, sold through a distribution channel, and priced so the issuer and the distributor both earn. The transcript compares it to putting sugar on a perfectly good plate of nuggets: technically possible, no reason to do it, and it ruins what you actually wanted.
The relevant Brazilian alternatives are straightforward. Tesouro Direto is the government platform for buying federal public bonds directly, and CDB is a bank deposit certificate that typically pays a percentage of the CDI interbank rate and may carry FGC deposit insurance up to a defined limit. A COE sits outside both of those simpler categories.
That is the structural point. A note is not automatically bad because it is complex. It is a problem when complexity is used to obscure how much of the return the seller keeps.
Why 9 in 10 structured notes lag Tesouro Direto
The transcript cites an analysis reporting that 9 in 10 COEs deliver an expected return below what the same money would have earned in Tesouro Direto. The cause is arithmetic, not bad luck: the commission and the capped upside are deducted before the investor sees the result.
A distributor who is paid on distribution will distribute what pays. The transcript puts the embedded commission on structured notes at 4% to 7%, and notes that a fund manager earns a slice of the administration fee, which is why a note tends to appear in a portfolio instead of a plain government bond.
Where the return goes, line by line:
| Component | Who captures it | Effect on the investor's return |
|---|---|---|
| Embedded commission on the COE | Distributor | 4% to 7% deducted up front |
| Administration fee on the fund that holds the note | Fund manager | A slice of the fee leaves the portfolio |
| Capped upside | Issuer | Any gain above the cap is forfeited |
| Coupon actually received | Investor | What is left after the items above |
| Tesouro Direto alternative | Investor | Captures the full CDI or inflation-linked rate |
The Brazilian market itself is now more transparent about this. Since 2023, distributors serving retail investors have had to disclose remuneration on investment products under the rules the CVM, Brazil's securities regulator, applies to distribution. That disclosure requirement is a reason the comparison the transcript makes is now easier to run on your own statements.
So the practical question is not whether the scenario is attractive. It is whether the payoff the investor actually receives, after the embedded commission, beats the CDI or an inflation-linked bond they could have bought in three clicks.
To run that comparison on your own statement:
- Find the disclosed remuneration on the COE under the CVM distribution rules in force since 2023, and subtract it from the gross return.
- Apply the note's cap to the underlying's performance to get the upside you can actually keep, not the upside in the marketing sheet.
- Compare the resulting net figure against the CDI or an inflation-linked bond available in three clicks.
If step 3 does not come out ahead, the arithmetic that produces the 9 in 10 result is working on you too.
The private-credit concentration problem
Concentration in private credit is the risk most retail portfolios underestimate. The portfolio in the transcript held more than 20 different credit issuers, which sounds diversified until you notice that all 20 sit on the same side of one trade: companies borrowing money in a high-interest-rate environment.
The transcript's central example is Raízen, the Brazilian sugar, ethanol and fuel joint venture. The video describes it as a company that entered the market in 2021 in an IPO and was valued in the tens of billions of reais, a profile an investor would not expect to be associated with a near-total loss of the credit. The specific valuation figures and the current status of Raízen's credit and equity are time-sensitive; check the company's own filings and B3, the Brazilian exchange, for the present state before acting on that example.
The mechanism is what matters. When you lend to 20 companies directly, one default can wipe out the accumulated excess return from the other 19. When you lend through a diversified credit fund or an incentivized infrastructure bond, the same macro environment applies but a single issuer no longer carries the whole outcome.
The other structural issue is maturity. The transcript notes assets maturing in 2040 and 2045 sitting inside a personal portfolio. Long-dated private credit means years of exposure to one company's ability to keep paying, and that is a very different risk from a government bond of the same duration.
| Route | Issuer count | Single-default effect | Maturity exposure |
|---|---|---|---|
| Direct lending to 20+ issuers | 20+ | One default can erase the excess return from the other 19 | Issuer-specific, including assets maturing 2040 and 2045 |
| Diversified credit fund | Spread across the fund | Diluted across many issuers | Set by the fund |
| Incentivized infrastructure bond | Per bond, often one project | Concentrated in that issuer | Tied to the project's own timeline |
| Government bond of the same duration | Single sovereign issuer | Sovereign rather than corporate credit | Matched duration, different risk |
Before relying on any of these routes, check the present state of a named issuer such as Raízen via its own filings and B3, since the valuation figures and credit status are time-sensitive.
Cash drag, and why idle money is not neutral
Money sitting in a checking account earns nothing for the account holder and earns the CDI overnight for the bank. The transcript describes exactly this: a cash balance set aside for investments, sitting uninvested while the screen shows the same number.
The mechanism is the float. When a deposit stays with the bank overnight, the bank can apply it at or near the CDI, and keeps the difference. Deposits in Brazil are covered by the FGC up to the limit set by the National Monetary Council, so an uninvested balance is protected but not remunerated.
The correction is trivial and requires no product decision. A daily-liquidity instrument such as a CDB at 100% of the CDI or a Tesouro Selic position earns while remaining available for the next investment. The cost of leaving it in the account is small per day and compounds over years.
How a 23-year-old founder's portfolio gets rebuilt
The rebuild in the transcript follows a simple split: aggressive in the operating company, conservative in the personal portfolio. The founder drew a salary of R$ 9.000 per month, had raised it from R$ 7.000 when he moved to São Paulo, and reinvested company profit rather than increasing his own withdrawal.
The consultancy's model is fee-based rather than commission-based. The transcript states that the consultancy cannot legally earn commission on distribution, which removes the incentive to place a high-fee product. That is a structural difference worth understanding before choosing any advisor: fee-only and commission-based advisory create different reasons to recommend the same asset.
| Dimension | Fee-only advisory | Commission-based advisory |
|---|---|---|
| How the advisor is paid | Fee paid by the client | Commission on distributed products |
| Stated legal constraint in the transcript | Cannot earn commission on distribution | Not described in the transcript |
| Resulting incentive on product choice | Removes the incentive to place a high-fee product | Not described in the transcript |
| Why the same asset may still be recommended | Different reasons to recommend, per the transcript | Different reasons to recommend, per the transcript |
The proposed allocation moved the personal portfolio toward roughly 70% to 80% fixed income across post-fixed, inflation-linked and pre-fixed sleeves, with the remainder in variable income, crypto exposure and an international slice. The transcript describes a simulation run over a three-year window, from July 2023 to July 2026, comparing the proposed portfolio against the CDI and the Brazilian stock index.
The rebuild can be read as an ordered sequence:
- Keep the personal withdrawal fixed at R$ 9.000 per month, after raising it from R$ 7.000 when he moved to São Paulo.
- Reinvest company profit rather than increasing that withdrawal.
- Move the personal portfolio toward roughly 70% to 80% fixed income, spread across post-fixed, inflation-linked and pre-fixed sleeves.
- Hold the remainder in variable income, crypto exposure and an international slice.
- Test the result over the three-year window from July 2023 to July 2026 against the CDI and the Brazilian stock index.
The stated rationale is that the founder already takes business risk, so the personal portfolio's job is to reduce the chance he has to sell the company at a bad moment. That logic is sound for the described profile, but the allocation percentages are the advisory team's proposal for one person's circumstances, not a template.
The rule of 72 does not change with portfolio size
Portfolio size does not meaningfully accelerate real doubling, and the transcript is explicit about that. The advisory team's illustration used a real return that doubles purchasing power roughly every seven years, and the host confirmed that the difference between R$ 1 million and R$ 100 million is access to a few closed deals, not a changed timeline.
The arithmetic behind the seven-year figure is the familiar rule of 72: at a 10% annual real return, money doubles in about 7.2 years. Change the assumption and the timeline moves with it. The transcript also frames the investor's simulation at a nominal 12% annual rate, so the real doubling period there depends on what inflation turns out to be, which nobody controls.
The useful takeaway is that the compounding math is not the bottleneck for a young investor. The contribution rate is, which is why the transcript returns repeatedly to the amount added each month rather than the percentage return on the existing balance.
What the CVM requires distributors to disclose now
Brazil's securities regulator has tightened what distributors must tell retail investors about their own compensation. The CVM rules on distribution remuneration in force since 2023 require disclosure of how the intermediary is paid on each product, which makes the commission argument in the transcript checkable against your own statement.
This matters because the transcript's strongest claim is about incentives, not about any single note. An assessor paid on distribution earns more from some products than others, and the products that pay more are not randomly distributed across the risk spectrum.
Readers can test the claim directly. Ask the person recommending the product for the total remuneration on it, in percentage terms, and compare that with a plain CDB or Tesouro bond. The comparison is now a disclosure question, not a speculative one.
None of this makes every structured note unsuitable. It makes the payoff formula and the compensation the two things to read first.
FAQ
- What is a COE in simple terms? A COE is a structured note issued by a bank whose return depends on a reference asset such as an index, a currency or a basket of stocks. The bank sets the payoff formula, and the note carries the issuer's credit risk, so it is not equivalent to a government bond or an insured deposit.
- Why do most COEs underperform Tesouro Direto? The transcript cites analysis reporting that 9 in 10 COEs deliver less than Tesouro Direto over their term. Embedded distribution commissions of roughly 4% to 7% and capped upside reduce the return before it reaches the investor, which is a structural drag rather than a market-timing problem.
- Is fee-based advisory better than commission-based advisory? Fee-based advisory removes the incentive to place products that pay the distributor more, which is the specific conflict the transcript highlights. It does not remove every conflict, and the quality of the recommendation still depends on the advisor's competence and the client's actual circumstances.
- Does investing internationally protect a Brazilian portfolio? A dollar-denominated allocation reduces exposure to real depreciation, which is the stated rationale in the transcript for a 15% international sleeve. Currency hedging does not remove market risk, and it introduces the tax and reporting obligations that apply to foreign holdings held by a Brazilian resident.
- How long does it take for money to double in real terms? At a 10% annual real return, purchasing power doubles in roughly seven years, which is the rule-of-72 approximation the transcript uses. A different assumed real return, or a period of high inflation, changes that timeline, and the number applies to the compounding math rather than to any specific investment.
Turning financial conversations into articles people can use
The value in a long portfolio review is rarely the final allocation. It is the reasoning that gets you there: why concentration in 20 credit issuers is different from diversification, why an uninvested balance is a cost rather than a neutral waiting position, why a fee-only model changes what gets recommended.
That reasoning usually already exists in a recorded conversation, and it tends to disappear once the video ends. If you have explanations, interviews or client walkthroughs sitting in a YouTube channel, Skala Blog can take the video, transcribe it, and produce a structured article from the same material, so the explanation keeps working after the recording stops.
Fork this article
Start a new branch from the same video, shaped your way. You keep the credit; the original keeps the attribution.
A fork in another language is filed as a translation of this article, so the two pages point at each other. You can unlink it later from the editor.
0/240
You are creating
- Format
- For
- Language
- Source
- Your angle
You will be asked to sign in before it is generated.
Buy credits