What Holding MLPs for 20 Years Actually Looks Like

The distributions compound. The tax deferral runs for over a decade. The basis resets at death. That's the thesis. This article is the math behind it: year-by-year basis erosion, the inflection when basis hits zero, the buy-more strategy that extends the deferral, the §751 shadow that grows while you hold, and what your heirs actually inherit under §1014. Every number is shown. Every assumption is stated.

This analysis models a specific approach: buying midstream MLP units in a taxable brokerage account, collecting distributions for decades, adding to positions when basis erodes, and holding through inheritance for a stepped-up basis. A multi-decade hold only works if the MLP's fundamentals hold up — check how to evaluate an MLP (coverage, DCF, leverage) before committing. Every number in this article is calculated under those assumptions.

In the first five years of holding a midstream MLP, you collect over $11,900 in cash distributions on a $37,500 investment and pay approximately $700 in federal tax. Your effective tax rate on cash received: 5.9%.

For comparison: $11,900 received as qualified dividends from a midstream C-corp ETF would cost approximately $2,240 in federal tax. The same cash flow taxed as ordinary income would cost approximately $4,270. The MLP structure delivered identical cash for $700.

Here are the base-case numbers. The scenario: 1,000 units of EPD purchased at $37.50/unit ($37,500 total). Current annualized distribution: $2.20/unit ($0.55/quarter). Current yield: approximately 5.9%.

After five years holding 1,000 units of a midstream MLP with 4% distribution growth and 80% return-of-capital rate, cumulative cash distributions total $11,917. Cumulative federal tax: $701.

All figures are nominal dollars. Distribution growth of approximately 4% roughly preserves purchasing power against long-run inflation. In real terms, the cash flow stream is approximately flat.

Around year 15–16 in this EPD scenario — earlier for higher-yield MLPs like MPLX — your original lot's basis hits zero. Every distribution after that point triggers immediate capital gain recognition under IRC §731(a), even though you haven't sold anything. The deferral machine stops for that lot. But it doesn't have to stop for your entire position.

Buying additional units creates fresh basis with its own independent deferral runway. This per-lot math is the core of the long-term MLP strategy.

Years 6–7: Basis Continues Declining

After Year 5, the pattern continues. Basis erodes at roughly 80% of each year's distribution. By Year 7, the original 1,000-unit lot has a basis of $23,599 — down 37% from the $37,500 purchase price.

At this pace — losing roughly $2,000–$2,300 of basis per year with the amount growing as distributions grow — Lot 1's basis reaches zero in Year 16. For the full mechanics of what happens when MLP basis reaches zero, see the dedicated guide. The summary: every distribution becomes taxable as long-term capital gain at 15% + 3.8% NIIT = 18.8%. Not catastrophic, but the 5.9% effective rate is over.

The Buy-More Strategy: Fresh Basis, Per Lot

This is the most technically critical concept in long-term MLP holding: basis is tracked per lot, not per position.

When you buy 200 additional units at $45/unit ($9,000) in Year 8:

• The 1,000 original units still have their own basis ($23,599 at start of Year 8). Their share of each quarterly distribution continues eroding their basis toward zero.
• The 200 new units have $9,000 of fresh basis. Their share of each quarterly distribution is absorbed as return of capital against this new basis — fully tax-deferred, just like your original lots in Year 1.
• You have NOT "reset the clock" on the original lots. You have added new lots with their own independent deferral runway. Each lot erodes on its own schedule.

Here's what the position looks like with two lots tracked independently. Lot 1 (1,000 units, original purchase) and Lot 2 (200 units, purchased at start of Year 8):

The Cascade: Extending the Deferral

The pattern continues. In Year 14, a third lot — 200 units at $50/unit ($10,000) — adds another layer of fresh basis. Here's what the three lots look like as Lot 1 approaches zero:

After 20 years holding 1,000 units of a midstream MLP with 4% distribution growth, buying 200 additional units in Years 8 and 14 when basis approached zero, the total cash distributions collected are $80,926. The total federal tax paid is $7,560. The effective tax rate on cash received is 9.3%.

For context: if the same $80,926 had been received as qualified dividends over 20 years, federal tax would have been approximately $15,200. As ordinary income, approximately $29,000. The MLP structure produced the same cash for $7,560.

The Decision That Defines the Strategy

The 20-year tape produces two very different tax outcomes depending on what happens at the end. This is the central decision of long-term MLP holding:

Time value of money note: A dollar of tax deferred 15 years is worth less than a dollar saved today. At a 5% discount rate, $3,000 in tax paid in Year 15 has a present value of approximately $1,450. The hold strategy backloads taxes (or eliminates them entirely at death), which means the present-value-adjusted tax burden is even lower than the nominal 9.3% suggests.

Holding direct MLPs means you have a filing obligation in every state where the partnership operates. EPD reports operations in approximately 27 states. ET in 44. In practice, most states have de minimis thresholds below which no filing is required, and many MLPs provide state-specific allocation worksheets. But depending on your positions, you may file 5–15 additional state returns per year.

Quantifying the cost over 20 years:

The AMLP investor files zero additional state returns. The midstream C-corp ETF investor files zero additional state returns. This is a real advantage of the ETF wrapper, and the comparison in Section 8 accounts for it.

For the full breakdown of which states require filing for each MLP, see the MLP state filing requirements guide.

Each year you hold, depreciation deductions flow through your K-1 and reduce your basis. These same deductions are simultaneously creating §751 "hot asset" exposure — the amount that would be recharacterized as ordinary income if you sold. The §751 amount grows silently every year you hold. You'll never see it on a statement. It materializes only on the Sales Schedule when you sell — or it vanishes entirely if you hold through inheritance.

For midstream MLPs with massive pipeline and processing infrastructure — EPD's depreciable asset base exceeds $60 billion — the §751 accumulation rate is approximately $1.50–$2.50 per unit per year. Here's what that looks like over the holding period:

After 20 years of holding, estimated §751 ordinary income exposure across all lots is approximately $36,000–$60,000. If sold, that's $10,600–$17,600 in additional federal tax — at ordinary income rates, not capital gains rates. This is the price of selling. The stepped-up basis at death is the reward for not paying it.

AMT note: For midstream MLPs, AMT preference items are typically minimal. Upstream MLPs with intangible drilling costs or percentage depletion can generate AMT exposure. If you hold non-midstream MLPs, check Box 15 of your K-1 for AMT items.

The numbers above assume everything goes right for 20 years. Here's what can go wrong. This section is what separates this article from every MLP cheerleading piece on the internet.

MLP-to-C-Corp Conversion

The biggest single risk to the hold-forever strategy is a forced taxable event you don't control. Multiple MLPs have converted to C-corp structure: Kinder Morgan (2014), ONEOK (2017), Targa Resources (2018), Williams (2018). A conversion typically triggers recognition of gain, including §751 recapture — the exact event this strategy is designed to avoid.

Evaluate conversion risk for any MLP you plan to hold for 20 years. EPD has structural characteristics that make conversion unlikely — its partnership agreement and the Duncan family's control structure favor MLP status. MPLX's dropdown relationship with Marathon Petroleum makes it more useful as an MLP than as a C-corp subsidiary. But no partnership agreement prevents conversion permanently.

Forced Disposition Through Merger or Acquisition

Your MLP can be acquired without your consent. Linn Energy went bankrupt. Crestwood was absorbed by Energy Transfer. Boardwalk was taken private. Over any 20-year window, a significant percentage of MLPs cease to exist. Most MLP mergers are taxable events for unitholders. Your "never sell" strategy can be forcibly terminated by the GP. When evaluating which MLPs to hold long-term, consider the sponsor's incentive to maintain the structure.

Distribution Cuts

Energy Transfer cut its distribution by 50% in 2020. What happens to the model?

Counterintuitively, a distribution cut slows basis erosion (less ROC reducing basis), which extends the deferral window. But it crushes cash flow. Modeling a 50% cut in Year 6 with recovery over three years: cumulative distributions fall by approximately $2,850, while cumulative tax paid drops by approximately $400–$600. The effective tax rate actually improves slightly. The real damage is to total return, not tax efficiency.

Phantom Income

If an MLP cuts its distribution but continues generating taxable K-1 income — which happens, because the partnership's operations generate income regardless of what they distribute — you can owe tax without receiving enough cash to pay it. This is rare for midstream but occurred with several MLPs during COVID. Over a 20-year holding period, the probability of encountering this at least once is non-trivial.

Legislative Risk to §1014

The entire endgame of this strategy rests on stepped-up basis at death under §1014. This provision has been targeted by multiple legislative proposals — in 2021, 2023, and as part of OBBBA negotiations. It is one of the most politically vulnerable provisions in the tax code.

If §1014 is eliminated or modified, the terminal tax benefit changes. However: the strategy still works well even without the step-up, because 20 years of tax-deferred compounding has standalone value. The effective tax rate on cash is 5.9% for the first 15 years regardless of what happens at death. The step-up is the cherry — the deferral is the sundae.

Concentration Risk

If you hold 3–4 MLPs for 20 years, you have significant sector concentration in North American energy midstream. The implicit bet: hydrocarbon volumes continue flowing through North American infrastructure for decades. That's probably a sound bet given LNG export growth, petrochemical demand, and the physical reality that pipelines are the lowest-cost transportation method. But state it as an assumption, not a certainty.

Tax Rate Changes

This model uses current federal rates throughout. Tax rates have changed multiple times in any 20-year period and will likely change again. If ordinary rates increase (e.g., from 37% to 39.6%), the value of tax deferral increases — making the hold strategy more advantageous, not less. If LTCG rates increase, the §731 gains after basis hits zero become more expensive, which strengthens the case for the buy-more strategy and for holding through inheritance.

The Hardest Part: Doing Nothing

Under IRC §1014, when an MLP investor dies, the heir's cost basis resets to fair market value on the date of death. Every dollar of basis erosion from decades of tax-deferred distributions — eliminated. Every dollar of §751 recapture exposure — eliminated. The heir starts fresh, as if they purchased the units at market price that day.

Here are the full numbers for the scenario modeled in this article:

Community Property States: The Double Step-Up

In community property states — Arizona, California, Idaho, Louisiana, Nevada, New Mexico, Texas, Washington, and Wisconsin — BOTH halves of community property receive a basis step-up when the first spouse dies, even the surviving spouse's half. In common law states, only the decedent's half steps up.

For a married couple in Texas holding $200,000 in MLPs as community property, the full $200,000 steps up — not just $100,000. This effectively doubles the step-up benefit compared to common-law states and is a significant advantage that most MLP articles overlook.

Direct MLPs produce the highest after-tax outcome under the hold-forever-and-inherit strategy. But the advantage comes with real costs. Here's the apples-to-apples comparison across three ways to access midstream energy cash flows.

All three columns assume the same $56,500 invested on the same schedule ($37,500 at Year 0, $9,000 at Year 8, $10,000 at Year 14), distributions collected as cash (not reinvested), and a 20-year holding period. Federal tax rates held constant.

AMLP assumptions: Effective yield ~4.5% after C-corp tax drag (vs. ~5.9% for direct MLP). Distribution growth ~3.5%. Tax characterization: ~60% qualified dividend (18.8% rate), ~40% ROC. NAV growth reduced by C-corp deferred tax liability accrual. Expense ratio 0.85%.

Midstream equity ETF assumptions: Invested in C-corps (ONEOK, WMB, KMI). Yield ~3.5% (all qualified dividends at 18.8%). Higher price growth (~4.5%/yr) due to lower distribution payout and retained earnings. Expense ratio ~0.35%.

For the full engine-computed side-by-side — corporate drag, the ETF’s own return-of-capital deferral, and the §1014 step-up on both sides, with every assumption editable — see the direct MLP vs ETF comparison.

I am not a registered investment advisor. What follows are the criteria I personally use when evaluating whether an MLP belongs in a multi-decade holding strategy. These are not recommendations to buy or sell any security.

1. Distribution Coverage Ratio

I look for 1.3x or higher to survive commodity cycles without cutting. At 1.8x, EPD has significant buffer. At 1.3x, MPLX has adequate cushion. Below 1.2x, I consider the cut risk too high for a 20-year hold.

2. Distribution Growth Trajectory

Over 20 years, the growth rate dominates total return. MPLX's 9.4% CAGR produces a radically different Year-20 distribution than EPD's 4.1%. Run both scenarios. The growth rate also determines how fast basis erodes — faster growth means more distributions, more ROC, and an earlier zero-basis inflection.

3. Balance Sheet Leverage

Debt-to-EBITDA below 4x. The MLP needs to exist in 20 years. Leverage kills partnerships during downturns. EPD runs at approximately 3.0x. MPLX at approximately 3.7x. I get cautious above 4x for a multi-decade hold.

4. Conversion Risk

Does the GP have an incentive to maintain the MLP structure? EPD's Duncan family benefits from the partnership. MPLX is more tax-efficient as an MLP dropdown vehicle for Marathon Petroleum. Targa and Williams converted because the C-corp structure served the business better. Evaluate the alignment between GP incentives and MLP structure before committing for 20 years.

5. Basis Erosion Rate

Higher ROC percentage means faster erosion means earlier §731 trigger means more frequent top-up purchases needed. This isn't inherently good or bad — it determines the cadence of your buy-more strategy. A 90% ROC MLP requires fresh capital sooner than a 75% ROC MLP.

6. Survival Probability

Has this partnership survived a commodity downturn? Does it have investment-grade credit? Is it large enough to be acquirer rather than target? Over 20 years, a significant percentage of MLPs will cease to exist through bankruptcy, merger, or conversion. The MLPs most likely to survive are the ones with the most essential infrastructure and the strongest balance sheets.

7. Capex Cycle and Debt Trajectory

Where is this MLP in its capital expenditure cycle? An MLP actively building — EPD with the Neches River terminal, MPLX with Permian and Marcellus expansions — is borrowing to fund growth. That borrowing increases your per-unit liability allocation under §752, which adds to your outside basis and extends the tax-deferral runway. An MLP that has finished its major projects and is deleveraging will shrink your liability allocation, accelerating basis erosion.

Neither is inherently good or bad — but it affects the timing of when you'll need to buy more units or start facing §731 gains. I look at the planned capex pipeline and the debt maturity schedule to understand where the liability allocation is likely headed over the next 5–10 years.

Current data for context — these are the MLPs I follow, not recommendations:

Life happens. The hold-forever thesis doesn't mean you're trapped. Here's the hierarchy of options if you need liquidity, ordered from most to least tax-efficient:

Having a plan for liquidity events is not a sign that the strategy is failing. It's a sign that you've thought beyond the spreadsheet.

I hold MLPs because the tax code rewards patient capital in ways that are hard to find elsewhere. The distributions compound, the deferral runs for decades, and the basis resets when it matters most. I've shown you the math, the risks, and the costs. The optionality — hold or pass on with no tax penalty at inheritance — is the real gift. What you do with it is up to you.

EPD has been publicly traded since 1998. A 20-year backtest — someone who bought in 2004 and held through 2024 — would show actual distributions received, actual basis erosion, actual unit appreciation. That data exists in 20 years of K-1 filings.

I have not yet reconstructed 20 years of K-1 data to validate this model against history, and I won't present unverified numbers. The model above uses current distribution rates projected forward with stated growth assumptions. When the historical reconstruction is complete, this section will be updated with the results.

This is more useful than a fabricated backtest. Real K-1 data captures the year-to-year variability in ROC rates, liability allocations, and income characterization that a smoothed model cannot. When published, it will either validate the model's assumptions or reveal where the smoothing created material divergence.