diff --git a/python/pylibcudf/pylibcudf/libcudf/fixed_point/__init__.pxd b/python/pylibcudf/pylibcudf/libcudf/fixed_point/__init__.pxd
new file mode 100644
index 00000000000..e69de29bb2d
diff --git a/python/pylibcudf/pylibcudf/libcudf/fixed_point/fixed_point.pxd b/python/pylibcudf/pylibcudf/libcudf/fixed_point/fixed_point.pxd
new file mode 100644
index 00000000000..e55574020f4
--- /dev/null
+++ b/python/pylibcudf/pylibcudf/libcudf/fixed_point/fixed_point.pxd
@@ -0,0 +1,8 @@
+# Copyright (c) 2021-2024, NVIDIA CORPORATION.
+
+from libc.stdint cimport int32_t
+
+
+cdef extern from "cudf/fixed_point/fixed_point.hpp" namespace "numeric" nogil:
+    cdef cppclass scale_type:
+        scale_type(int32_t)
diff --git a/python/pylibcudf/pylibcudf/libcudf/scalar/scalar.pxd b/python/pylibcudf/pylibcudf/libcudf/scalar/scalar.pxd
index 4b40a8a26f6..a51413669c5 100644
--- a/python/pylibcudf/pylibcudf/libcudf/scalar/scalar.pxd
+++ b/python/pylibcudf/pylibcudf/libcudf/scalar/scalar.pxd
@@ -4,9 +4,9 @@ from libc.stdint cimport int32_t, int64_t
 from libcpp cimport bool
 from libcpp.string cimport string
 from pylibcudf.libcudf.column.column_view cimport column_view
+from pylibcudf.libcudf.fixed_point.fixed_point cimport scale_type
 from pylibcudf.libcudf.table.table_view cimport table_view
 from pylibcudf.libcudf.types cimport data_type
-from pylibcudf.libcudf.wrappers.decimals cimport scale_type
 
 
 cdef extern from "cudf/scalar/scalar.hpp" namespace "cudf" nogil:
diff --git a/python/pylibcudf/pylibcudf/libcudf/wrappers/decimals.pxd b/python/pylibcudf/pylibcudf/libcudf/wrappers/decimals.pxd
deleted file mode 100644
index 558299501d6..00000000000
--- a/python/pylibcudf/pylibcudf/libcudf/wrappers/decimals.pxd
+++ /dev/null
@@ -1,16 +0,0 @@
-# Copyright (c) 2021-2024, NVIDIA CORPORATION.
-
-from libc.stdint cimport int32_t, int64_t
-from pylibcudf.libcudf.types cimport int128
-
-
-cdef extern from "cudf/fixed_point/fixed_point.hpp" namespace "numeric" nogil:
-    # cython type stub to help resolve to numeric::decimal64
-    ctypedef int64_t decimal64
-    # cython type stub to help resolve to numeric::decimal32
-    ctypedef int64_t decimal32
-    # cython type stub to help resolve to numeric::decimal128
-    ctypedef int128 decimal128
-
-    cdef cppclass scale_type:
-        scale_type(int32_t)