EPS (Earnings Per Share) is a widely used metric to assess a company’s financial performance. The calculation of EPS is the net earnings of a company divided by the number of outstanding shares. However, EPS can be calculated using Generally Accepted Accounting Principles (GAAP) or non-GAAP measures.

GAAP EPS follows the accounting rules every US public company must apply, so it includes all expenses and gains the standards require. Non-GAAP EPS, often called adjusted EPS, starts from the same net income and removes items the company considers outside its ongoing operations, such as stock-based compensation, amortization of acquired intangibles or restructuring charges.

The gap between the two can be small or very large. Over the last four reported quarters, Apple’s reported EPS and its GAAP EPS were almost identical. Pfizer’s reported EPS added up to 3.05, while its GAAP EPS added up to 0.76. This guide explains both measures and shows how to pull and compare them with the EODHD Fundamentals API in Python.

Register & Get Data

GAAP EPS: definition and formula

GAAP is a set of accounting standards that public companies in the United States must follow. These standards are designed to ensure consistency and transparency in financial reporting, so GAAP measures provide a standardized way to compare the financial performance of different companies.

The formula is:

GAAP EPS = (net income – preferred dividends) / weighted average shares outstanding

Companies report two versions. Basic EPS uses the weighted average of shares actually outstanding. Diluted EPS also counts the shares that options, warrants, convertible securities and restricted stock units could add, so it is equal to or lower than basic EPS when the company is profitable. When analysts and data providers say “GAAP EPS” without a qualifier, they usually mean diluted EPS.

Non-GAAP EPS: what companies adjust

Non-GAAP EPS starts from GAAP net income and adds back or removes items that management considers not representative of ongoing operations. The most common adjustments are:

  • stock-based compensation, a large expense for many technology companies
  • amortization of intangible assets acquired in takeovers
  • restructuring, severance and impairment charges
  • gains or losses on investments and asset sales
  • one-off legal, tax or acquisition-related items, and the tax effect of all the above

Each company chooses its own adjustments, so non-GAAP EPS is not standardized across companies. Used appropriately, it can show the trend of the underlying business. It can also make results look better than they are. For that reason, the Securities and Exchange Commission (SEC) requires companies that publish non-GAAP measures to present the closest GAAP measure with equal prominence, provide a reconciliation between the two, and explain why the non-GAAP measure is useful.

GAAP vs non-GAAP EPS at a glance

GAAP EPSNon-GAAP (adjusted) EPS
RulesUS accounting standardsChosen by each company
Comparable across companiesYesOnly with care
Includes stock-based compensationYesUsually excluded
Includes one-off charges and gainsYesUsually excluded
Where it appearsIncome statement, SEC filingsEarnings release, with a reconciliation
Typical useTrailing P/E, statutory comparisonsEarnings beats and misses against analyst estimates

Neither measure is the “true” one. GAAP EPS is the consistent baseline. Non-GAAP EPS is the company’s view of its operating performance, and it is the number earnings headlines usually compare with analyst estimates.

How to get your EODHD API key

1. Use the “demo” API key to test our data from a limited set of tickers without registering:
AAPL.US | TSLA.US | VTI.US | AMZN.US | BTC-USD | EUR-USD
Real-Time Data and all of the APIs (except Bulk) are included without API calls limitations with these tickers and the demo key.
2. Register to get your free API key (limited to 20 API calls per day) with access to:
End-Of-Day Historical Data with only the past year for any ticker, and the List of tickers per Exchange.
3. To unlock your API key, we recommend to choose the subscription plan which covers your needs.

Register & Get Data

Fundamentals data is part of the Fundamentals Data Feed and All-In-One plans, and each fundamentals request costs 10 API calls. The demo key returns fundamentals for the demo tickers, so you can run the Apple examples below without registering.

Where EPS lives in the EODHD Fundamentals API

The Fundamentals API (https://eodhd.com/api/fundamentals/{TICKER}) returns EPS in three places:

SectionFieldWhat it contains
Earnings::HistoryepsActual, epsEstimate, epsDifference, surprisePercentEPS for each reported quarter, the analyst estimate and the surprise
HighlightsDilutedEpsTTM, EarningsShareDiluted EPS over the trailing twelve months
Financials::Income_StatementnetIncome, netIncomeApplicableToCommonSharesThe net income behind GAAP EPS, by quarter and year

According to EODHD, Earnings::History provides non-GAAP earnings, and when a company does not publish non-GAAP earnings, the figure is the GAAP one. The comparison below is consistent with that and shows what it means in practice.

With the filter parameter you can request a single section or field instead of the full JSON:

import os
import requests

API_KEY = os.environ.get("EODHD_API_KEY", "demo")
url = "https://eodhd.com/api/fundamentals/AAPL.US"
params = {"api_token": API_KEY, "fmt": "json", "filter": "Earnings::History"}
history = requests.get(url, params=params, timeout=60).json()
print(history["2026-06-30"])
{'reportDate': '2026-07-30', 'date': '2026-06-30', 'beforeAfterMarket': 'AfterMarket', 'currency': 'USD', 'epsActual': 2.02, 'epsEstimate': 1.88, 'epsDifference': 0.14, 'surprisePercent': 7.4468}

The same call with filter=Highlights::DilutedEpsTTM returns a single number instead of a JSON object.

Reported vs GAAP EPS for four companies

To see how far apart the two measures can be, this script takes the last four reported quarters for each company and puts the reported EPS from Earnings::History next to a GAAP EPS rebuilt from the income statement (net income attributable to common shareholders, or net income when that field is empty, divided by the shares outstanding at quarter end):

import os
import sys

import pandas as pd
import requests

API_KEY = os.environ.get("EODHD_API_KEY", "demo")


def fundamentals(ticker):
url = f"https://eodhd.com/api/fundamentals/{ticker}"
resp = requests.get(url, params={"api_token": API_KEY, "fmt": "json"}, timeout=120)
resp.raise_for_status()
return resp.json()


def eps_table(ticker, quarters=4):
data = fundamentals(ticker)
history = data["Earnings"]["History"]
income = data["Financials"]["Income_Statement"]["quarterly"]
balance = data["Financials"]["Balance_Sheet"]["quarterly"]

rows = []
for quarter in sorted(history, reverse=True):
h = history[quarter]
if h.get("epsActual") is None or quarter not in income or quarter not in balance:
continue
net = float(income[quarter]["netIncomeApplicableToCommonShares"] or income[quarter]["netIncome"])
shares = float(balance[quarter]["commonStockSharesOutstanding"])
rows.append({
"ticker": ticker,
"quarter": quarter,
"reported_eps": h["epsActual"],
"estimate": h["epsEstimate"],
"gaap_eps_rebuilt": round(net / shares, 2),
})
if len(rows) == quarters:
break
df = pd.DataFrame(rows)
df["gap"] = (df["reported_eps"] - df["gaap_eps_rebuilt"]).round(2)
ttm = {
"ticker": ticker,
"reported_ttm": round(df["reported_eps"].sum(), 2),
"gaap_rebuilt_ttm": round(df["gaap_eps_rebuilt"].sum(), 2),
}
return df, ttm


if __name__ == "__main__":
tables, ttms = [], []
for t in sys.argv[1:]:
df, ttm = eps_table(t)
tables.append(df)
ttms.append(ttm)
print(pd.concat(tables).to_string(index=False), "\n")
print(pd.DataFrame(ttms).to_string(index=False))

I ran it on November 3, 2026 with python gaap_vs_reported.py AAPL.US CRM.US PFE.US INTC.US:

 ticker    quarter  reported_eps  estimate  gaap_eps_rebuilt   gap
AAPL.US 2026-09-30 2.05 1.98 2.10 -0.05
AAPL.US 2026-06-30 2.02 1.88 2.02 0.00
AAPL.US 2026-03-31 2.01 1.94 2.01 0.00
AAPL.US 2025-12-31 2.84 2.67 2.84 0.00
CRM.US 2026-07-31 4.95 2.35 4.29 0.66
CRM.US 2026-04-30 3.88 3.13 2.42 1.46
CRM.US 2026-01-31 3.81 3.05 2.07 1.74
CRM.US 2025-10-31 3.25 2.86 2.17 1.08
PFE.US 2026-06-30 0.77 0.68 -0.04 0.81
PFE.US 2026-03-31 0.75 0.72 0.47 0.28
PFE.US 2025-12-31 0.66 0.57 -0.29 0.95
PFE.US 2025-09-30 0.87 0.63 0.62 0.25
INTC.US 2026-09-30 0.52 0.28 0.63 -0.11
INTC.US 2026-06-30 0.30 0.10 -2.16 2.46
INTC.US 2026-03-31 0.29 0.01 -0.73 1.02
INTC.US 2025-12-31 0.15 0.08 -0.12 0.27

ticker reported_ttm gaap_rebuilt_ttm
AAPL.US 8.92 8.97
CRM.US 15.89 10.95
PFE.US 3.05 0.76
INTC.US 1.26 -2.38

The Apple rows run with the demo key. Salesforce, Pfizer and Intel need your own key with fundamentals access.

What the comparison shows

Apple’s reported EPS and its rebuilt GAAP EPS are equal in three of the four quarters. The 0.05 difference in the September quarter is small enough to come from the share count used in the rebuild (see the next section).

For the other three companies, the reported EPS in Earnings::History is the adjusted figure, and the gap is large:

  • Salesforce: 15.89 reported over four quarters against 10.95 GAAP. The reported figure is higher in every quarter.
  • Pfizer: 3.05 reported against 0.76 GAAP, with two quarters in which GAAP EPS was negative while reported EPS stayed positive.
  • Intel: 1.26 reported against a GAAP loss of 2.38 per share. In the June 2026 quarter alone, reported EPS was 0.30 and GAAP EPS was -2.16. Adjustments can also go the other way: in the September quarter, the rebuilt GAAP EPS (0.63) was higher than the reported 0.52.

In practice:

  • Use epsActual, epsEstimate and surprisePercent from Earnings::History to study earnings surprises. The estimate and the actual are on the same basis, which is the basis earnings headlines use.
  • Use DilutedEpsTTM, or net income from the income statement, when you need GAAP earnings, for example to compare profitability across companies or to compute a P/E on reported profits.
  • When the two differ a lot, the company’s reconciliation in its earnings release explains which items make up the gap.

Where the rebuilt figure is approximate

The rebuilt GAAP EPS divides quarterly net income by the shares outstanding at the end of the quarter. Companies divide by the weighted average diluted shares over the quarter, so the rebuilt value can differ by a few cents, especially for companies that buy back or issue many shares. The September 2026 quarter for Apple (2.10 rebuilt) is an example.

Two other details are worth handling in your own code. netIncomeApplicableToCommonShares can be empty, so the script falls back to netIncome. And fiscal quarters don’t always end with the calendar quarter: Salesforce’s quarters end in January, April, July and October, so match income statement and balance sheet entries by their own dates rather than by calendar quarter.

The comparison covers four companies and four quarters. It shows how much the two measures can diverge, not how often they do across the market.

Conclusion

GAAP EPS is the standardized measure of what a company earned for each share. Non-GAAP EPS is the company’s adjusted view, and it is the figure analyst estimates and earnings surprises are built on. For some companies they are the same number. For others, like Pfizer and Intel in the last four quarters, they tell very different stories.

The EODHD Fundamentals API gives you both sides: reported EPS, estimates and surprises in Earnings::History, and GAAP earnings in Highlights and the income statement. Try the Apple example with the demo key, then run the comparison on the companies you follow.

For the other numbers in the Fundamentals API, see Fundamentals: Basics Explained, and for the full list of fields, the Fundamentals API documentation. Upcoming report dates are available through the Earnings Calendar API.

Register & Get Data

FAQs

Q: What is GAAP EPS?

GAAP EPS is earnings per share calculated under US Generally Accepted Accounting Principles: net income minus preferred dividends, divided by the weighted average shares outstanding. Diluted GAAP EPS also counts shares from options, convertibles and restricted stock units.

Q: What is non-GAAP earnings per share?

Non-GAAP EPS is an adjusted EPS that a company publishes alongside its GAAP figure. It removes items such as stock-based compensation, amortization of acquired intangibles, restructuring charges and one-off gains or losses. Each company chooses its own adjustments and must reconcile them to GAAP EPS.

Q: What is the difference between GAAP and non-GAAP EPS?

GAAP EPS includes everything the accounting standards require. Non-GAAP EPS excludes items the company considers non-recurring or non-operating. In the data above, the gap over four quarters ranged from almost zero for Apple to 4.94 per share for Salesforce.

Q: Is EPS in earnings reports GAAP or non-GAAP?

Earnings headlines usually compare the company’s adjusted (non-GAAP) EPS with analyst estimates, which are prepared on the same basis. According to EODHD, Earnings::History contains non-GAAP EPS when the company publishes it, and GAAP EPS otherwise.

Q: Which EPS should I use for the P/E ratio?

A P/E on GAAP EPS reflects all reported costs and is comparable across companies. A P/E on adjusted EPS (often called an adjusted or operating P/E) is lower when adjustments are large. Whichever you use, use the same basis for every company you compare.

Do you enjoy our articles?

We can send new ones right to your email box