diff --git a/openbb_platform/extensions/quantitative/openbb_quantitative/helpers.py b/openbb_platform/extensions/quantitative/openbb_quantitative/helpers.py
index 1ecb1e8693e5..ff478d40366d 100644
--- a/openbb_platform/extensions/quantitative/openbb_quantitative/helpers.py
+++ b/openbb_platform/extensions/quantitative/openbb_quantitative/helpers.py
@@ -78,3 +78,94 @@ def validate_window(input_data: Union["Series", "DataFrame"], window: int) -> No
         raise ValueError(
             f"Window '{window}' is greater than the input data length '{len(input_data)}'"
         )
+
+
+def deflated_sharpe_stats(
+    input_data: "Series",
+    trials: int,
+    rfr: float = 0.0,
+    trials_sr_std: Union[float, None] = None,
+) -> dict:
+    """Compute the deflated Sharpe ratio of a return series.
+
+    The deflated Sharpe ratio (Bailey & Lopez de Prado, 2014) is the
+    probability that the true Sharpe exceeds the expected maximum Sharpe of
+    `trials` zero-skill strategies — i.e. whether a result selected as the
+    best of `trials` attempts reflects skill rather than selection. Adjusts
+    for sample length and return skewness/kurtosis.
+
+    Parameters
+    ----------
+    input_data : Series
+        Return series (per-period returns, not prices).
+    trials : int
+        Number of strategy variants tried before selecting this one.
+    rfr : float, optional
+        Per-period risk-free rate, by default 0.0.
+    trials_sr_std : float, optional
+        Standard deviation of the per-period Sharpe estimates across the
+        trials. Defaults to the null 1/sqrt(n-1) when unknown.
+
+    Returns
+    -------
+    dict
+        sharpe (per-period), expected_max_sharpe, deflated_sharpe_ratio,
+        observations, trials.
+
+    Raises
+    ------
+    ValueError
+        If `trials` < 1 or the series has fewer than 3 observations or zero
+        variance.
+    """
+    # pylint: disable=import-outside-toplevel
+    from numpy import asarray, e as np_e, finfo, sqrt
+    from scipy.stats import norm
+
+    if trials < 1:
+        raise ValueError("trials must be >= 1")
+    returns = asarray(input_data, dtype=float)
+    n = len(returns)
+    if n < 3:
+        raise ValueError("need at least 3 observations")
+    mu = returns.mean() - rfr
+    sd = returns.std()  # population, consistent with the reference implementation
+    # `sd == 0` is exact and a constant series does not reach it: its standard
+    # deviation is floating-point residue rather than a true zero, so a flat 0.1%
+    # series has sd ~1e-19 and divides out to a Sharpe of ~1e16 -- finite, and past
+    # every isfinite guard after this one. Deflating that returned 1.0, i.e. certainty
+    # of a real edge, for the one input carrying no information about one. Compare
+    # against the resolution of a float at the scale of the data instead, so a
+    # genuinely low-volatility series still gets a number.
+    # The residue grows with the number of terms summed, so the floor is n eps rather
+    # than eps: measured at most 1.96 eps x scale over constant series spanning values
+    # 1e-7..1e3 and lengths 3..10000, while a real series with sigma=1e-12 sits more
+    # than ten orders of magnitude above n eps x scale.
+    if not sd > n * finfo(float).eps * abs(returns).max():
+        raise ValueError("zero-variance returns")
+    sr = mu / sd
+    skew_ = (((returns - returns.mean()) / sd) ** 3).mean()
+    kurt_ = (((returns - returns.mean()) / sd) ** 4).mean()  # non-excess, normal = 3
+
+    if trials_sr_std is None:
+        trials_sr_std = 1.0 / sqrt(n - 1)
+    if trials == 1:
+        bar = 0.0
+    else:
+        euler = 0.5772156649015329
+        z1 = norm.ppf(1 - 1.0 / trials)
+        z2 = norm.ppf(1 - 1.0 / (trials * np_e))
+        bar = trials_sr_std * ((1 - euler) * z1 + euler * z2)
+
+    denom = 1 - skew_ * sr + (kurt_ - 1) / 4 * sr**2
+    if denom <= 0:
+        raise ValueError("degenerate return moments")
+    dsr = norm.cdf((sr - bar) * sqrt(n - 1) / sqrt(denom))
+
+    return {
+        "sharpe": float(sr),
+        "expected_max_sharpe": float(bar),
+        "deflated_sharpe_ratio": float(dsr),
+        "observations": n,
+        "trials": trials,
+    }
diff --git a/openbb_platform/extensions/quantitative/openbb_quantitative/models.py b/openbb_platform/extensions/quantitative/openbb_quantitative/models.py
index 5082739d3159..f0e4782bebb7 100644
--- a/openbb_platform/extensions/quantitative/openbb_quantitative/models.py
+++ b/openbb_platform/extensions/quantitative/openbb_quantitative/models.py
@@ -48,6 +48,16 @@ class OmegaModel(BaseModel):
     omega: float
 
 
+class DeflatedSharpeModel(BaseModel):
+    """Deflated Sharpe ratio model (Bailey & Lopez de Prado, 2014)."""
+
+    sharpe: float
+    expected_max_sharpe: float
+    deflated_sharpe_ratio: float
+    observations: int
+    trials: int
+
+
 class SummaryModel(BaseModel):
     """Summary model."""
 
diff --git a/openbb_platform/extensions/quantitative/openbb_quantitative/performance/performance_router.py b/openbb_platform/extensions/quantitative/openbb_quantitative/performance/performance_router.py
index d2659c3f14b1..9ec55694d0b4 100644
--- a/openbb_platform/extensions/quantitative/openbb_quantitative/performance/performance_router.py
+++ b/openbb_platform/extensions/quantitative/openbb_quantitative/performance/performance_router.py
@@ -9,6 +9,7 @@
 from openbb_core.app.router import Router
 from openbb_core.provider.abstract.data import Data
 from openbb_quantitative.models import (
+    DeflatedSharpeModel,
     OmegaModel,
 )
 from pydantic import PositiveInt
@@ -269,3 +270,72 @@ def sortino_ratio(
     results_ = df_to_basemodel(results)
 
     return OBBject(results=results_)
+
+
+@router.command(
+    methods=["POST"],
+    examples=[
+        PythonEx(
+            description="Get the Deflated Sharpe Ratio of a strategy selected as the best of 100 variants.",
+            code=[
+                'stock_data = obb.equity.price.historical(symbol="TSLA", start_date="2023-01-01", provider="fmp").to_df()',  # noqa: E501
+                'returns = stock_data["close"].pct_change().dropna()',
+                'obb.quantitative.performance.deflated_sharpe_ratio(data=returns, target="close", trials=100)',
+            ],
+        ),
+        APIEx(
+            parameters={
+                "target": "close",
+                "trials": 100,
+                "data": APIEx.mock_data(
+                    "timeseries",
+                    sample={"date": "2023-01-01", "close": 0.05},
+                ),
+            },
+        ),
+    ],
+)
+def deflated_sharpe_ratio(
+    data: list[Data],
+    target: str,
+    trials: PositiveInt,
+    rfr: float = 0.0,
+) -> OBBject[list[DeflatedSharpeModel]]:
+    """Get the Deflated Sharpe Ratio.
+
+    The Deflated Sharpe Ratio (Bailey & Lopez de Prado, 2014) answers the question a raw Sharpe cannot: does this
+    result reflect skill, or is it what the luckiest of `trials` attempts would show by chance? When a strategy is
+    selected as the best of many tested variants, the expected maximum Sharpe under pure noise grows with the number
+    of trials — so the winning backtest must be scored against that bar, adjusting for sample length and the skewness
+    and kurtosis of returns. A value near 1 means the Sharpe survives the number of attempts made; a value near 0.5 or
+    below means the result is indistinguishable from selection luck. Essential whenever parameters were optimized or
+    multiple strategies compared before reporting a result.
+
+    Parameters
+    ----------
+    data : list[Data]
+        Time series data of per-period returns.
+    target : str
+        Target column name.
+    trials : PositiveInt
+        Number of strategy variants tried before selecting this one.
+    rfr : float, optional
+        Per-period risk-free rate, by default 0.0
+
+    Returns
+    -------
+    OBBject[list[DeflatedSharpeModel]]
+        Sharpe, expected max Sharpe of `trials` zero-skill attempts, and the deflated Sharpe ratio.
+    """
+    # pylint: disable=import-outside-toplevel
+    from openbb_core.app.utils import (
+        basemodel_to_df,
+        get_target_column,
+    )
+    from openbb_quantitative.helpers import deflated_sharpe_stats
+
+    df = basemodel_to_df(data)
+    series_target = get_target_column(df, target)
+    stats_ = deflated_sharpe_stats(series_target, trials=trials, rfr=rfr)
+
+    return OBBject(results=[DeflatedSharpeModel(**stats_)])
diff --git a/openbb_platform/extensions/quantitative/tests/test_deflated_sharpe.py b/openbb_platform/extensions/quantitative/tests/test_deflated_sharpe.py
new file mode 100644
index 000000000000..d81bf1c4a864
--- /dev/null
+++ b/openbb_platform/extensions/quantitative/tests/test_deflated_sharpe.py
@@ -0,0 +1,80 @@
+"""Test the deflated Sharpe ratio helper."""
+
+import math
+
+import pandas as pd
+import pytest
+from extensions.quantitative.openbb_quantitative.helpers import (
+    deflated_sharpe_stats,
+)
+
+# Reference values computed with the numguard reference implementation of
+# Bailey & Lopez de Prado (2014) on the same deterministic series
+# (population moments, null trial dispersion 1/sqrt(n-1)).
+RETURNS = pd.Series(
+    [0.003 + 0.01 * math.sin(i) + 0.004 * math.cos(3 * i) for i in range(90)]
+)
+
+
+def test_matches_reference_implementation():
+    """Parity against the reference implementation."""
+    result = deflated_sharpe_stats(RETURNS, trials=50)
+    assert result["sharpe"] == pytest.approx(0.4101142526095976, abs=1e-9)
+    assert result["expected_max_sharpe"] == pytest.approx(
+        0.24128764537744862, abs=1e-9
+    )
+    assert result["deflated_sharpe_ratio"] == pytest.approx(
+        0.9399085246020616, abs=1e-7
+    )
+    assert result["observations"] == 90
+    assert result["trials"] == 50
+
+
+def test_single_trial_has_no_deflation():
+    """With one trial the bar is zero: pure probabilistic Sharpe vs 0."""
+    result = deflated_sharpe_stats(RETURNS, trials=1)
+    assert result["expected_max_sharpe"] == 0.0
+    assert result["deflated_sharpe_ratio"] == pytest.approx(
+        0.9999199954582253, abs=1e-7
+    )
+
+
+def test_deflation_is_monotonic_in_trials():
+    """More attempts must never make the same result more credible."""
+    values = [
+        deflated_sharpe_stats(RETURNS, trials=n)["deflated_sharpe_ratio"]
+        for n in (1, 10, 100, 1000)
+    ]
+    assert all(values[i] > values[i + 1] for i in range(len(values) - 1))
+
+
+def test_invalid_inputs_raise():
+    """Trials and sample-size guards."""
+    with pytest.raises(ValueError):
+        deflated_sharpe_stats(RETURNS, trials=0)
+    with pytest.raises(ValueError):
+        deflated_sharpe_stats(RETURNS.head(2), trials=10)
+    with pytest.raises(ValueError):
+        deflated_sharpe_stats(pd.Series([0.01] * 10), trials=10)
+
+
+def test_zero_dispersion_raises_across_values_and_lengths():
+    """A constant series carries no information about an edge, at any scale.
+
+    `sd == 0` is exact and a constant series does not reach it: its standard
+    deviation is floating-point residue rather than a true zero, so the ratio came
+    out finite (~1e16) and reached the deflation arithmetic, which answered 1.0.
+    The residue depends on both the value and the length, so this is checked over a
+    grid: a guard calibrated on a single series passes while still leaking elsewhere.
+    """
+    for value in (1e-7, 1e-4, 0.001, 0.01, 1.0, 100.0):
+        for n in (3, 10, 250, 5000):
+            with pytest.raises(ValueError, match="zero-variance"):
+                deflated_sharpe_stats(pd.Series([value] * n), trials=4)
+
+    # The guard is relative to the scale of the data: a real but very quiet series
+    # still gets a number.
+    import numpy as np
+
+    quiet = pd.Series(np.random.default_rng(1).normal(0, 1e-8, 250))
+    assert 0 <= deflated_sharpe_stats(quiet, trials=4)["deflated_sharpe_ratio"] <= 1