# The Certificate Payback Teardown, Three Programs to a Break-Even Month

*Turn any training program's completer outcomes, local entry wages, and your current pay into a risk-adjusted gain and a dated break-even month you can defend.*

- Canonical URL: https://www.refolk.ai/candidates/guides/certificate-payback-break-even-teardown
- Pillar: Reading the market
- Format: Teardown
- Published: 2026-11-21
- Last reviewed: 2026-11-21
- Reading time: 23 min
- Keywords: is a certificate program worth it for a career change, how long to pay back a trade school certificate, starting salary after certificate program, compare training programs by graduate earnings, entry level wage for career changers by occupation, retraining return on investment calculator

## Key takeaways

- Price a certificate's starting pay with the local OEWS 10th and 25th percentiles, not the median: the national HVAC median of $61,010 sits $20,960 above the 10th percentile.
- Rank training programs on completion rate times second-quarter employment rate before earnings: Savannah Tech's HVAC program out-earns Atlantic Tech's but its multiplier is 0.047 against 0.297, six times worse.
- A program's Q2 median times four is a ceiling, not a salary: it counts every employed exiter, completers or not, including people who worked only part of the quarter.
- With tuition covered and 14 weeks of wages lost against a $22,590 counterfactual, the strongest HVAC row in the trace recovers its cost in about 7.7 months of work, while the weakest needs 203.
- Date the break-even month from the day you file for funding, adding the 2 to 8 week training-account wait, the training weeks, and a job-search quarter.
- In Refolk's index, adding an entry-level filter to 4,435 US HVAC Technician profiles returned the same 4,435, so profile counts cannot stand in for entry pay in this trade.

This trace prices a short certificate before you pay for it. It asks whether the pay in the new occupation covers the tuition plus the wages you give up while training, and in which month it catches up. It is written for career changers holding a tuition quote, and it follows one worked case all the way through: an HVAC certificate (SOC 49-9021) priced against three real program rows, from public records to a go, cheaper-provider, or no-go call. The wrong turns stay in, because that is where programs start to look better than they are.

## What does a retraining payback trace produce?

For each program you are considering, the trace produces three numbers: a first-year pay range in the new occupation, a risk-adjusted annual gain over what you earn now, and a dated break-even month. Each one traces to a public record you can cite.

Terms used throughout:

- **SOC code**: the federal Standard Occupational Classification code for the job you are training into. HVAC mechanics and installers are 49-9021.
- **OEWS percentiles**: wage points from the Occupational Employment and Wage Statistics survey. P10 is the wage that 10% of workers earn less than, and P25 is the wage that a quarter earn less than.
- **Q2 median earnings**: from the federal training-provider index. This is the median total earnings in the second quarter after exit, across everyone from the program who was employed then, whether their training was publicly funded or not.
- **Completion rate**: the share of a program's exiters who finished.
- **Q2 employment rate**: the share of exiters employed in that second quarter.
- **Risk multiplier**: completion rate times Q2 employment rate. The trace uses it as a rough estimate of your odds of finishing and working.
- **Net cost**: the tuition left after training accounts and grants, plus the wages you lose while training.
- **Break-even month**: the calendar month in which your cumulative risk-adjusted gain equals your net cost, counted from the day you file for funding.

### The counterfactual pay in this trace

You need your own current pay, and the worked case needs a stand-in. I use $22,590 a year, which is 150% of the poverty line for a single person at the 2024 guideline. The Department of Education subtracts the same counterfactual in its Workforce Pell value-added earnings test, so it is a public, defensible baseline. When you run your own case, replace it with your gross annual pay. If you earn more than $22,590, every gain in this guide shrinks.

## The payback trace, step by step

The procedure runs from the wage row to the call in eight steps. The first three are about two hours of desk research and the rest is arithmetic. The time estimates are mine and have not been measured.

#### Retraining payback trace

1. **Pin the SOC code and local wage row** - Find the SOC code for the target occupation and pull its OEWS row for the metro where you will work. Record P10, P25, median, the release month and the area.
2. **Pull candidate programs from the index** - On TrainingProviderResults.gov, record exiter count, completion, Q2 employment, Q2 median, cost and length for each program you can attend. Treat any -1 as missing, never as zero.
3. **Cross-check the Scorecard row** - If the provider takes federal student aid, note its College Scorecard 1-year completer median next to the index figure, with the cohort caveats.
4. **Set three first-year pay scenarios** - Downside is local P10, base is the lower of Q2 times four or local P25, and upside is local P25. If Q2 times four is below P10, it becomes the downside too.
5. **Discount by completion times employment** - Multiply completion by Q2 employment to get the risk multiplier, then apply it to the gain over your current pay.
6. **Net the cost** - Subtract your training account and any Pell from published tuition, add the wages lost while training, and put the 2 to 8 week funding wait on the timeline.
7. **Date the break-even month** - Divide net cost by the monthly risk-adjusted gain, then count forward from filing through the funding wait, the training and a job-search quarter.
8. **Make the call** - Decide go, cheaper provider, or no-go, with one line per number naming the record it came from.

One fork. Some practitioners run step 6 before step 2, because eligibility for a training account (an Individual Training Account, or ITA, the voucher a local workforce board issues under WIOA, the Workforce Innovation and Opportunity Act) decides which programs are candidates at all. If your board funds only programs on its own eligible list, pull that list first and run step 2 only on what appears on it.

## Which wage counts as starting pay for a career changer?

Use the local P10 as your downside and the local P25 as the ceiling for your base case, both from the same OEWS release and the same metro area. The median is not starting pay, and a national row is not your local row.

Here is how far the HVAC anchors sit apart, and how far they moved between releases:

| Point | May 2022 national | May 2025 national | Change |
|---|---|---|---|
| P10 | $36,170 | $40,050 | +10.7% |
| P25 | $44,100 | $48,360 | +9.7% |
| Median | $51,390 | $61,010 | +18.7% |

Two things follow. First, the May 2025 national median of $61,010 is 52% above P10, or $20,960 a year more than the 10th-percentile wage. Second, entry rows move: P10 rose 10.7% between these two releases, so an old row understates today's floor. The May 2025 values come from secondary mirrors of the BLS release, so verify them on bls.gov before you cite them.

The worked case uses the Miami metro row from CareerOneStop, taken from the May 2024 release:

- Local P10: $38,270
- Local P25: $46,710

Miami's P10 is below the national May 2025 P10 of $40,050. Because the two figures come from different releases, that gap mixes area and time, which is why step 1 records the release month as well as the area.

> **Note:** Sources disagree on which percentile is "entry"
>
> One state labor-market office defines entry, upper entry and experienced wages as the 10th, 25th and 75th percentiles, while a secondary source describes the entry wage as roughly the 25th. CareerOneStop itself cautions that the 10th percentile is not necessarily a starting wage, and BLS has not published which percentile a new entrant should expect, so price both and call it a range.

> **Rule:** Never price starting pay off the median
>
> The median describes a worker halfway up the occupation's experience ladder, not someone holding a fresh certificate. Your base case may never exceed the local P25 from the same release and area as your P10.

### Wrong turn 1: the national median

Suppose the career changer in the trace had used the national median. The gain over $22,590 would be $38,420 a year, or $3,202 a month, which recovers $6,082 of lost wages in 1.9 months with no discount applied. Most answers to "is a certificate program worth it for a career change" stop at a number like that. It is four times faster than the most honest figure for the best program in this trace.

## How do you compare training programs by graduate earnings?

Read five fields from the training-provider index for each program: exiter count, completion, Q2 employment, Q2 median and cost. Then multiply completion by Q2 employment and rank on that product before you look at earnings. Ranking on earnings puts programs in a different order from ranking on completion, and completion decides whether you get the earnings at all.

The index (TrainingProviderResults.gov, with its Data Dictionary) publishes a bulk file of programs on states' Eligible Training Provider lists. Each row gives provider, program, credential, SOC codes, completion rate, Q2 employment rate, Q2 median earnings, cost and length. Employment and earnings cover all students, funded or not. Each release reflects the state lists as of a cutoff date and covers a multi-year window of exiters, so check the About page for the window you are reading.

Here are the three HVAC rows in the trace:

| Program | Completion | Q2 employed | Q2 median | Q2×4 | Multiplier |
|---|---|---|---|---|---|
| Hennepin Tech HVAC | 46% | 85% | $12,066 | $48,264 | 0.391 |
| Atlantic Tech HVAC/R | 69% | 43% | $6,299 | $25,196 | 0.297 |
| Savannah Tech AC | 7% | 67% | $7,555 | $30,220 | 0.047 |

The rows are from program year 2023. The last two columns are my arithmetic. In your own trace, keep only programs you can actually attend. I keep all three here because they are the three shapes you will run into.

**0.047 - Savannah Tech AC risk multiplier, 7% completion times 67% Q2 employment**

Its Q2 median beats Atlantic Tech's, but its odds of finishing and working are six times worse (0.047 against 0.297).

### Wrong turn 2: ranking on the Q2 median

On earnings alone, Savannah's $7,555 beats Atlantic's $6,299. On the multiplier, Savannah falls to last by a wide margin. Students who drop out never appear in the earnings figure, but they still paid tuition. A high employment rate paired with low completion is the classic false positive.

### Why Q2 times four is a cap, not a salary

The Q2 median covers every exiter employed in that quarter, completers or not, including people who worked only part of it. That is why the trace uses Q2×4 only to cap the base case. One useful agreement shows up here: Hennepin's Q2×4 of $48,264 sits within $100 of the national P25 of $48,360. When a program's own record lands on P25 like that, it supports P25 as the base case and P10 as the downside.

The fork in step 4 applies to two of the three rows. Atlantic's $25,196 and Savannah's $30,220 are both below Miami's P10 of $38,270. When that happens, the program's own record says its working exiters are not reaching the local floor, so Q2×4 becomes both the base case and the downside, and the local P25 stays as the upside.

> **Watch out:** A -1 is not a zero, and not a pass
>
> The index writes -1 or null where DOL suppressed a value: a cohort too small to protect privacy, no data reported, or a significant quality issue. Read the exiter count before any rate, and do not let a blank cell drop out of your comparison as if it were neutral.

### Cross-checking with College Scorecard

If the provider takes federal student aid, its College Scorecard field-of-study row gives you a second opinion based on a different cohort. Scorecard measures students who completed a credential and were federally aided, counting only those employed and not enrolled. Its earnings are W-2 wages plus positive self-employment earnings. It now reports measurement points at 1 to 5 years after completion, along with a field-of-study national median. One secondary source puts its suppression threshold at under 30 students, but I have not verified that against ED's own documentation. Nobody has publicly established how far Scorecard figures differ from the index for the same program, so record both and do not average them.

## What does the certificate actually cost after aid?

Net cost is published tuition, minus your training account and any Pell, plus the wages you lose while training. The funding wait goes on the timeline as well, because a sourced example says it usually takes 2 to 8 weeks to get training paid for.

Training-account caps are set locally and vary widely:

| Area | ITA cap | Note |
|---|---|---|
| New Jersey | Up to $7,500 | Recommended limit; Pell and state funds must be used first |
| Tennessee boards | $2,500 to $8,000 | Varies by board |
| Most Texas boards | $4,000 to $20,000 | Varies by board |

No national figure exists for net tuition after Workforce Pell plus a training account. To qualify for Workforce Pell, a program must be 150 to 599 clock hours and run at least eight but less than 15 weeks. I could not verify the award amount for programs of that length, so ask the school's financial aid office for the figure in writing.

Sector matters before aid does. Georgetown's Center on Education and the Workforce found that public institutions' net cost is roughly one-third of the cost at for-profit institutions. In TICAS's reading of 1-year Scorecard data, the typical graduate makes less than $30,000 in 88% of for-profit certificate programs, 78% of private non-profit ones, and 46% of public ones.

### Lost wages in the trace

At $22,590 a year, training full time costs about $434 a week in forgone pay:

- 8 weeks: $3,475
- 14 weeks: $6,082

For the trace I assume a 14-week full-time program, the longest whole-week length inside the Workforce Pell window, with tuition fully covered by a training account. Net cost is then $6,082 of lost wages. Read the real length off the index row. If you keep some paid hours while you train, subtract what you keep.

## Worked case: three HVAC programs dated to a break-even month

With tuition covered and 14 weeks of wages lost, the strongest row (Hennepin) recovers its cost in about 7.7 months of work, which lands in month 16 after filing for funding. Atlantic needs about 94 months of work and Savannah about 203.

The headline number shrinks at each honest step:

#### Hennepin HVAC, from headline pay to expected gain

| Stage | Figure | Note |
| --- | --- | --- |
| National median | $61,010 | the number most answers quote |
| Program Q2 times four | $48,264 | a cap, not a salary |
| Local P25 base case | $46,710 | Miami metro, the lower of the two |
| Gain over $22,590 | $24,120 | before any discount |
| Risk-adjusted gain | $9,431 | times 0.391 |

*The same certificate shrinks from a $61,010 headline to a $9,431 expected annual gain once each wrong turn is removed.*

Here is the same arithmetic for all three rows, with the median as a fourth row for contrast:

| Case | Monthly risk-adjusted gain | Months per $1,000 | Months to recover $6,082 |
|---|---|---|---|
| Hennepin, base | $786 | 1.3 | 7.7 |
| Atlantic, base | $64.50 | 15.5 | 94 |
| Savannah, base | $29.90 | 33.4 | 203 |
| Median, no discount (wrong) | $3,202 | 0.3 | 1.9 |

The "months per $1,000" column is the one to carry into your own case. Every $1,000 of tuition your training account does not cover adds that many months.

The other scenarios:

- **Hennepin downside (P10):** $38,270 minus $22,590 is $15,680, times 0.391 gives $511 a month, so 11.9 months to recover. The upside equals the base, because the P25 cap binds.
- **Atlantic upside (P25):** $597 a month gives 10.2 months. Reaching it means landing in-field at the local P25, and the program's own record of $25,196 annualised says most of its working exiters do not.
- **Savannah upside (P25):** $94.50 a month gives about 64 months, even on the generous assumption.

### Wrong turn 3: skipping the discount

Without the multiplier, Savannah's base gain of $7,630 a year recovers $6,082 in 9.6 months, which looks almost as good as Hennepin. Discounted, it takes 203 months. Nothing about the wage changed. All of that difference comes from the 93% of students who do not finish.

### Putting a date on it

#### Hennepin base case, filing to break-even

1. **File for funding** - wait 2 to 8 weeks for a training account
2. **Train** - 14 weeks full time, $6,082 of wages forgone
3. **Search** - one quarter, matching the Q2 measure
4. **Earn** - $786 a month risk-adjusted
5. **Break even** - month 16 from filing, month 20 on the downside

*Break-even is a calendar month counted from filing, not from the first paycheck.*

Count it out. The funding wait (up to 8 weeks), training (14 weeks) and a search quarter (13 weeks) add up to 35 weeks, about 8 months before the first paycheck. Add 7.7 months and the base case breaks even in month 16 after filing. Add 11.9 months instead and the downside lands in month 20. If you file in March, the base case breaks even in July of the following year.

I budget a full quarter for the job search because the index measures employment in the second quarter after exit, so the program's own employment rate already includes months of searching. If you assume you start work the week you finish, you are counting on months the data never promised.

### Sanity check against the literature

ED cites research putting the earnings gains from short-term programs at $1,200 to $2,000 per year. At those gains, $6,082 of lost wages takes 36 to 61 months to recover. The Hennepin base case of $9,431 a year sits well above that range, mostly because the stand-in current pay is low. If you earn well above $22,590, expect your trace to land closer to the literature than to this case.

**Payback trace worksheet**

```
Target SOC code / title:
Wage area and OEWS release month:
Local P10 (downside):            source:
Local P25 (upside / base cap):   source:
Program, provider, program year:
Exiter count (read first):
Completion rate:
Q2 employment rate:
Q2 median earnings:   x4 =
Base first-year pay = lower of (Q2 x4, local P25):
Downside = local P10, or Q2 x4 if lower:
Risk multiplier = completion x Q2 employment:
Current gross annual pay:
Annual gain (base) = base pay - current pay:
Risk-adjusted monthly gain = annual gain x multiplier / 12:
Published tuition:
Minus training account:          Minus Pell:
Plus lost wages = weekly pay x training weeks:
Net cost:
Months to recover = net cost / risk-adjusted monthly gain:
Funding wait (weeks) + training (weeks) + search (13 weeks):
Break-even month (base / downside):
Call (go / cheaper provider / no-go) and the one-line reason:
```

*Fill one copy per program; every line needs its source and release month.*

## How does a payback trace go wrong?

Almost every error in this trace makes the program look better than it is. The common ones are annualising Q2 pay, using the median, using a stale or national row, ignoring people who don't finish, misreading suppressed cells, treating the Pell test as payback, counting any job as in-field, and borrowing a cohort that does not match you.

1. **Q2×4 as annual pay.** *Looks like:* a clean salary figure. *What it hides:* it covers every employed exiter, completers or not, including people who worked only part of the quarter. *Check:* Savannah's $30,220 rests on a program where 7% complete. Use it as a cap only.
2. **Median as starting pay.** *Looks like:* the headline wage for the occupation. *What it hides:* the national HVAC median overstates entry pay by $20,960 against P10. *Check:* use the local P10 and P25 from one release.
3. **Stale or national rows.** *Looks like:* an official BLS number. *What it hides:* the May 2022 P10 is 10.7% below May 2025, and Miami's P10 differs from the national one. *Check:* write down the release month and the area before using any figure.
4. **Ignoring non-completers.** *Looks like:* a high employment rate. *What it hides:* drop-outs pay tuition but never show up in earnings. *Check:* multiply completion by employment. Savannah's 67% falls to 0.047.
5. **Suppressed cells read as zero or as good.** *Looks like:* a blank, or a program with no bad numbers. *What it hides:* a -1 means a small cohort, no data reported, or a quality flag. Suppression below 10 students hides 22% of participants, so the visible set is tilted toward large providers. *Check:* read the exiter count first. No data does not mean no risk.
6. **Treating Workforce Pell value-added earnings (VAE) as your payback.** *Looks like:* a federal seal of value. *What it hides:* the test is published tuition ≤ adjusted median earnings minus 150% of the poverty line, using only working completers three years out, adjusted for regional prices, with non-workers excluded. *Check:* read VAE as a legal ceiling on tuition, not as an expected gain.
7. **Employment that is not in-field.** *Looks like:* a strong Q2 employment rate. *What it hides:* the rate counts any job. Georgetown found that only 44% of certificate holders work in their field, and those who do earn 37% more than those who don't. Out-of-field holders earn just 1% more than high-school graduates. *Check:* ask the provider for training-related placement.
8. **Scorecard cohort mismatch.** *Looks like:* a second confirming source. *What it hides:* Scorecard counts only federally aided completers who are not still enrolled, and a WIOA-funded adult may not resemble that cohort. *Check:* record the cohort definition next to the figure.

> Earnings figures describe the people who found work. Your tuition bill is owed whether or not you are one of them.

One test that does not work as a shortcut: Workforce Pell's outcome floors (70% completion within 150% of normal time, and 70% of completers employed in Q2). Taken at face value, none of the three HVAC rows would clear both. But the denominators differ, because Pell's employment floor counts completers while the index rate counts all exiters. So treat that as a reason to ask questions, not as a verdict on eligibility.

**Questions to send the provider before enrolling**

```
Before I enroll in [program name], could you send the following for your most recent cohort?
1. How many students started, and how many completed within the normal program length?
2. Of completers, how many were working in a job related to the training two quarters after finishing?
3. What was the typical starting hourly wage in those training-related jobs, and how was it collected?
4. The full published cost, including tools, books and exam fees, and the program length in weeks and clock hours.
5. Is the program on the state Eligible Training Provider list, and which local workforce boards have funded students in it?
Thank you. I am comparing programs on completion and training-related placement, so the cohort counts matter as much as the rates.
```

*Send to admissions or career services; ask for figures in writing with the cohort year.*

## What can profile data tell you about entry pay?

For a trade, not much. In Refolk's index of professional profiles, adding an entry-level seniority filter to a US search for "HVAC Technician" returned exactly the same 4,435 profiles as the unfiltered search, so the index offers no usable seniority split for this occupation.

**4,435 - US profiles titled HVAC Technician in Refolk's index**

The count was identical with and without an entry-level filter, so trades profiles rarely carry a usable seniority signal.

| Market | Profiles | Sample with HVAC titles | Top employer |
|---|---|---|---|
| United States | 4,435 | at least 17/25 (68%) | ColonialWebb |
| Canada | 490 | 22/25 (88%) | Howe Sound Pulp and Paper |

What this proves: the title exists at scale, US profiles outnumber Canadian ones 9.1 to 1, and the US sample names real local employers. The top names were ColonialWebb (3), the US Air Force (2), Morris-Jenkins, and Howell's Heating & Air.

Where it lies: the 4,435 US profiles equal only 1.1% of OEWS employment for the occupation, and the sample's regions cluster in Charlotte and Raleigh. A count like that is a skewed slice, not a census, and it carries no pay at all. Use it to find employers whose entry postings you can read against your P10 to P25 band. Public wage percentiles remain the only defensible anchor for entry pay.

## Making the call and keeping it current

Make the call on two variables: the risk multiplier, and where the program's own pay record sits against the local wage band. Then check that both the base and downside break-even months fall inside the period you can fund.

#### Go, cheaper provider, or no-go

Horizontal axis runs from Low completion x employment to High completion x employment. Vertical axis runs from Q2 x4 below local P10 to Q2 x4 at or above local P25.

| Quadrant | What it means |
| --- | --- |
| Good pay, poor odds | Look for a better-finishing or cheaper provider for the same SOC |
| Go | Proceed if base and downside break-even fit your runway; Hennepin's row |
| No-go | Walk away; the Atlantic and Savannah rows sit here |
| Places people, below the floor | Ask for training-related placement before paying |

*A program needs both decent odds of finishing and working and a pay record at or above the local floor.*

How the three rows resolve:

- **Go:** Hennepin's shape. Its multiplier is the best of the three, its own pay record reaches local P25, and it breaks even in month 16 (base) or month 20 (downside).
- **Cheaper provider:** when the occupation pays but the program's cost or multiplier does not. Price a public provider first, given the one-third net-cost finding.
- **No-go:** Atlantic and Savannah as they stand. Their Q2×4 figures are below local P10, and they take 94 and 203 months to recover only the lost wages.

The whole trace assumes in-field work, and the gain disappears if you land outside the field. Before paying tuition, it is worth checking how much of your current history actually transfers. [Refolk](/candidates) scores how well your resume fits a specific posting, so running it against a handful of local entry-level HVAC postings shows which of your existing experience counts and which gaps the certificate has to close.

### What to re-check, and when

- **OEWS:** every release carries a reference month. Re-pull P10 and P25 when a new one appears for your area.
- **The index:** each release covers a new window of exiters as of a state-list cutoff. Re-read the program row, starting with the exiter count.
- **ITA caps:** these are set by local boards and change. Confirm the cap with your board before step 6.
- **Workforce Pell VAE:** ED publishes it no later than three months before the award year begins. When per-program figures first appear has not been publicly established, so check the Federal Register notice.

The search quarter is the one part of the timeline you directly control, and every week cut from it moves the break-even month earlier. Once the certificate is done, Refolk can rewrite your resume from your own history with the new credential included, tailor it to each posting, and draft the cover letter, so applications go out in the first week rather than the sixth.

#### Before you sign the enrollment form

- [ ] The SOC code, wage area and OEWS release month are written next to P10 and P25.
- [ ] No figure in the trace is a median or a national row standing in for local entry pay.
- [ ] Each program's exiter count was read before its rates, and no -1 was treated as a zero.
- [ ] Q2 times four is used only as a cap on the base case.
- [ ] Each program has a risk multiplier, and ranking was done on it before earnings.
- [ ] Net cost includes lost wages for the real program length, not just tuition.
- [ ] The training-account cap and any Pell amount are confirmed in writing by the board or the school.
- [ ] Base and downside break-even are calendar months counted from filing, including the funding wait and a search quarter.
- [ ] The provider has answered in writing on training-related placement.
- [ ] The call names the record behind every number.

## Frequently asked questions

### Is a certificate program worth it for a career change?

It is worth it when the program's completion rate times its employment rate, applied to the gap between local entry pay and your current pay, recovers tuition plus lost wages within a period you can fund. In the HVAC trace here, one program broke even in about 7.7 months of work and another needed 203 months on the same assumptions. That spread comes from completion, not from the trade itself.

### What starting salary should I expect after a certificate program?

Plan between the local 10th and 25th percentile wage for the target occupation, from the same OEWS release and metro. Use the program's Q2 median times four only as a ceiling. CareerOneStop warns that the 10th percentile is not necessarily a starting wage, and sources disagree on whether the entry figure means the 10th or 25th percentile, so keep both as your range.

### How long does it take to pay back a trade school certificate?

Divide net cost (tuition after aid plus lost wages) by the monthly risk-adjusted gain, then add the funding wait, the training weeks and a job-search quarter. For the strongest HVAC row in this trace, recovering $6,082 of lost wages took about 7.7 months of work, which puts break-even around month 16 after filing. The literature ED cites suggests 36 to 61 months for typical gains.

### Why can't I just multiply the program's quarterly earnings by four?

Because the Q2 median counts every exiter employed in that quarter, whether they completed or not, including people who worked only part of it. Multiplying by four can overstate pay. It can also understate it. Treat it as a cap on your base case, and never as a salary forecast. Savannah Tech's $30,220 annualised figure comes from a program where only 7% of students complete.

### What does a -1 mean in the training provider data?

It means the value was suppressed. The cohort may have been too small to protect privacy, no data may have been reported, or DOL may have found significant quality issues. It is never a zero and never a sign that things went well. Read the exiter count before any rate, and ask the provider directly. Programs with fewer than 10 students are suppressed, and they hold 22 percent of participants.

### Does Workforce Pell eligibility mean a program pays off?

No. The value-added earnings test caps tuition by comparing it with completers' earnings three years out, minus 150% of the poverty line. It counts only working completers and adjusts for regional prices. It drops the people who never finish or never work, so a program can pass and still take an individual years to break even.

---

*From the Refolk guide library. I revise these guides rather than replacing them, so the current version is always at https://www.refolk.ai/candidates/guides/certificate-payback-break-even-teardown*
